Critique and recommended actions from domain experts.
1 Expert: Medical Device Regulatory Affairs Director (Class III MDR & Notified Body Strategy)
Knowledge: EU MDR Class III device approval, notified-body engagement, clinical investigation authorisation, post-market surveillance, AI Act high-risk medical device compliance, international regulatory harmonisation
Why: The plan's success depends critically on securing notified-body alignment and CE marking by Month 132 (2037-10-31), with slippage risk to Month 144-156 (2038-10-31 to 2039-10-31). The expert reviews the Evidence and Regulatory Strategy decision, the AI Act change-control framework (EUR 2M budget), the preliminary QMS audit trigger (Month 18 = 2028-04-30), and scientific-advice sequencing. They validate whether the evidence package (chronic animal data, first-in-human design, decoder version-lock) meets notified-body expectations and whether thresholds set before data (e.g., 25% signal loss kill gate) are defensible under MDR scrutiny.
What: Audit the regulatory milestones (notified-body engagement Month 6 = 2027-04-30, QMS Month 9 = 2027-07-31, scientific-advice request Month 12 = 2027-10-31, preliminary audit Month 18 = 2028-04-30, architecture lock Month 18 = 2028-04-30) and validate the AI Act compliance roadmap. Confirm that decoder change-control rules align with post-CE modification thresholds. Identify unmet notified-body dialogue on pivotal-trial evidence design before Month 18 (2028-04-30) architecture lock, so rework does not slip CE marking.
Skills: Notified-body liaison and dossier management, MDR annex compliance, clinical-investigation protocol design under ISO 14155, AI Act risk classification and change-control governance, quality-management documentation, regulatory timeline and risk mapping
Search: MDR Class III regulatory affairs, notified body engagement medical devices, AI Act high-risk device compliance, clinical investigation authorisation
1.1 Primary Actions
- By Month 1 (2026-11-30): Draft and send written partnership-scope documents to imec, KU Leuven, UZ Leuven, Ghent/Antwerp, UMC Utrecht and CEA-Leti/Clinatec specifying committed resources, facility hours, staff FTEs, patient-access commitments and decision deadlines. Request signed LOIs from all six by Month 3 (2027-01-31). Simultaneously, publish a 'Partner Decline Decision Tree' showing which fallback activates first, lead-time per fallback, and the board decision point if activation slips past Month 4 (2027-02-28).
- By Month 2 (2026-12-31): Name at least three anchor investors or investor consortia (strategic, venture, public) and publish their written selection criteria for G1 commitment at Month 12 (2027-10-31). Define what 'committed in writing' means per tranche (e.g., 'signed term sheet, milestone triggers, investor board seat'). Publish a 'Gate Decision Authority and Timeline' document assigning binding veto power to a named board entity or independent trustee; commit to written verdicts within 5 business days of each gate (G1, Foundry, G3, G4).
- By Month 3 (2027-01-31): Commission an independent health-economic and eligible-population analysis covering ALS incidence and eligible-pool size in EU/EEA (current assumption: 500–1,400 per year), willingness-to-pay from three HTA bodies, spinal cord injury eligible pool, centre capacity per indication, and payer-acceptance thresholds. If ALS eligible pool is <800 per year and payers reject communication-only outcomes, pre-register a 'Pivot Gate' at Month 24 (2028-10-31) to decide spinal cord injury or accept a subsidy model.
- By Month 3 (2027-01-31): Establish an independent consent advocate (clinical ethicist, patient-representative background) by contract; recruit or form a Data Monitoring Committee (five+ members: two external statisticians, one patient representative, one ethicist, one external clinician) with binding authority to pause or stop implantation. Pre-register quantitative safety kill gates before any implantation: SAE thresholds (≥2 in first 10 implants or ≥3 in first 20 trigger pause), chronic-signal-loss thresholds (>25% in 12 months triggers pause), bandwidth floor (below 5 bits/s sustained in ≥50% at Month 72 = 2032-10-31 triggers review). Publish gates to ethics committees and notified body.
- By Month 6 (2027-04-30): If foundry commitments from 4+ member states are not in writing, execute the dual-sourcing decision immediately: contract two EU chip suppliers and two thin-film substrate suppliers with pre-priced qualification by Month 30 (2029-04-30). Document reallocation: foundry failure triggers automatic +EUR 35M draw on T1 and CE marking shift to Month 138 (2038-04-30) (estimate). Establish the patient-protection trust with EUR 5M floor, held outside investor reach, with independent trustees; fund it separately from operating capital.
- By Month 12 (2027-10-31): Close T0 (EUR 10–15M) in writing and confirm founder seed of at least EUR 5M from Month 3 (2027-01-31). Publish a written verdict ('proceed', 'defer', 'switch') within 5 business days of Month 12 (2027-10-31), signed by the board and notified to all investors, the Commission and fallback partners. If G1 is missed, do not renegotiate; instead, automatically defer Months 13–24 (2027-11-30 to 2028-10-31) spend and execute the 'defer-or-stop' fallback plan (which must be written before Month 12 = 2027-10-31). Request Commission scientific advice on clinical investigation and AI Act scope with notified-body feedback attached.
1.2 Secondary Actions
- By Month 1 (2026-11-30): Establish the design-history file (DHF) in a secure, versioned repository (password-protected shared drive or validated electronic system); assign a DHF custodian (QA Manager or delegate) and document the DHF governance before any design work begins. Initiate an upstream-dependency register listing all tools, software, materials and processes with country-of-origin and export-control status; publish a draft by Month 6 (2027-04-30).
- By Month 2 (2026-12-31): Hire or contract a Quality Assurance Manager with ISO 13485 and MDR experience. Establish a change-control process charter (document control, training, supplier audits, non-conformance logs, internal audits) and begin ISO 13485 system setup. By Month 9 (2027-07-31), request a preliminary QMS audit from the selected notified body.
- By Month 3 (2027-01-31): Commission a Leuven climate and infrastructure resilience assessment covering multi-day heatwaves above 35°C, drought, flooding, power reliability, cooling and water for manufacturing and compute. Install backup power and redundant cooling by Month 18 (2028-04-30); replicate EU-hosted compute at a second EU site; design the wearable for at least 12 hours battery with passive cooling and throttling, tested at 35–40°C.
- By Month 6 (2027-04-30): Engage a notified body (e.g., TÜV Süd, BSI, DEKRA Assure) for a pre-submission meeting; present architecture options (thin-film surface, penetrating cortical, endovascular), planned preclinical programme, first-in-human plan and decoder version-lock strategy. Request written feedback on evidence gaps and regulatory timeline.
- By Month 9 (2027-07-31): Commission an AI Act compliance assessment from external legal counsel (biotech or medical-device AI focus) covering decoder classification as high-risk, required documentation, training-data provenance, model-validation protocols and change-control thresholds. Deliver a written compliance roadmap by Month 9 (2027-07-31); budget EUR 2M for AI Act documentation, audit and administration over 120 months.
- By Month 12 (2027-10-31): Publish the upstream-dependency register with mitigation plans per component. Sign MOUs with two chip and two substrate suppliers specifying qualification milestones (Month 30 = 2029-04-30), price, supply commitment and export-control status. Initiate supplier audits and dual-source qualification.
- By Month 18 (2028-04-30): Ratify the architecture (thin-film surface primary, penetrating cortical fallback) based on FTO, surgical-risk, chronic-stability and bandwidth scoring from preclinical and published human data. Conduct freedom-to-operate analysis for each architecture and publish blocking-patent findings. Fund pilot-line-to-validated-manufacturing transfer with specific cost and schedule before design lock. Request preliminary QMS audit feedback from the notified body.
- By Month 24 (2028-10-31): Pre-register the first-in-human clinical investigation plan (5–10 participants, coordinated multinational consent, DMC oversight, 12-month follow-up) and submit to ethics committees for approval. Publish the final feasibility verdict (six-part: scientific, engineering, clinical, regulatory, supply-chain, financing) with evidence summary and named contingencies.
1.3 Follow Up Consultation
Recommend next consultation at Month 1 (2026-11-30) to review partner-outreach templates, anchor-investor criteria and gate decision-authority documents before sending. At Month 6 (2027-04-30), review partner LOI status and foundry commitment signals; finalize notified-body engagement and DMC charter. At Month 12 (2027-10-31), review T0 close status and gate verdict drafting; confirm scientific-advice request. At Month 18 (2028-04-30), review architecture ratification and FTO results; assess supplier MOUs and QMS audit findings. At Month 24 (2028-10-31), review feasibility-verdict drafting and indication-pivot gate criteria; confirm payer-acceptance and eligible-population analysis results. Key decision points: Month 3 (2027-01-31) (partner LOIs), Month 12 (2027-10-31) (T0 and G1), Month 18 (2028-04-30) (architecture and foundry switch decision), Month 24 (2028-10-31) (feasibility verdict and indication pivot gate).
1.4.A Issue - Financing structure is aspirational, not credible; gates lack enforcement mechanisms and fallback trigger clarity.
The plan assumes EUR 0 at Month 0 (2026-10-31) and EUR 2.5B committed over tranches T0–T4, but specifies no investor commitment letters, return expectations, or cost-of-capital assumptions. Gate G1 (Month 12 = 2027-10-31) requires EUR 10–15M 'committed in writing', yet no underwriter or anchor investor is named. If G1 fails, the plan says 'stop, defer or switch' but does not specify: who decides? How many days to decision? What is the alternative 'switch' pathway? The IPCEI foundry (EUR 350M, four+ member states) is hoped-for; if it fails at Month 18 (2028-04-30), dual-sourcing adds EUR 35M and 6–12 months, which compresses CE marking and strains T1–T2, yet no reallocation rule is pre-registered. The result is that gates are advisory, not binding.
1.4.B Tags
- financing_commitment_gap
- gate_enforcement_undefined
- founder_capital_burden
- ipcei_dependency_undeclared
1.4.C Mitigation
By Month 1 (2026-11-30): name at least three anchor investors (strategic, venture or public) and publish their written criteria for G1 commitment (e.g., 'EUR 10M on architecture lock + FTO clear + notified-body pre-submission held'). By Month 3 (2027-01-31): publish a 'Gate Decision Authority and Timeline' document naming the board, independent trustees, or an investor syndicate with binding veto power; commit to a written verdict within 5 business days of each gate. By Month 6 (2027-04-30): if foundry commitments from four+ member states are not in writing, execute the dual-sourcing decision immediately (do not defer to Month 18 = 2028-04-30). Reallocation: if foundry fails, T1 automatic +EUR 35M draw, shifting T2 start by 6 months and CE marking to Month 138 (2038-04-30) (estimate); document this in the tranche schedule before Month 12 (2027-10-31).
1.4.D Consequence
Without named investors and binding gate authority, a missed funding gate will trigger 3–6 months of board negotiation, scope renegotiation and delay. During this gap, the team cannot commit spend, cannot hire, and will lose founders and early staff to more-certain competitors. The result: project attrition, broken partner relationships (Leuven or fallback sites), and eventual abandonment or a defaulting of critical milestones (architecture lock, FTO, consent safeguards) into a compressed Months 24–36 (2028-10-31 to 2029-10-31) window, raising rework and failure risk.
1.4.E Root Cause
The plan treats financing gates as thresholds for good governance but does not operationalise them as hard stops. Founders are asked to raise money without pre-agreed investor personas, return structures or syndicate roles. Decisions are assumed to occur 'by consensus' without naming who has authority, so gates become aspirational targets rather than binding decision points. The foundry dependency is listed but not locked, so its failure is treated as a 'switch' rather than an automatic trigger with pre-costed alternatives.
1.5.A Issue - The 5,000-implant target is not credibly achievable and is mis-stated as a 'miss' rather than a false requirement; the plan does not account for the commercial-viability consequence.
The plan states the 5,000-implant target is 'missed in all scenarios' and reports 1,050 cumulative implants at Month 180 (2041-10-31) (central case). Yet it also reports revenue EUR 45M per year (300 implants/year × EUR 150,000) against opex EUR 111M per year, an operating gap of EUR 66M/year and break-even at 1,125 implants/year. The plan does not state: is the company profitable at 1,050 implants? (Answer: no, it is EUR 7.5M short per year.) Does the EUR 2.5B envelope cover the EUR 66M/year gap through Month 180 (2041-10-31)? (Calculation: 36 months at EUR 66M/year = EUR 198M; the plan has EUR 985M in T4 (Months 132–180 = 2037-10-31 to 2041-10-31), so yes, barely, but the company never reaches positive cash flow and survives only via continued investor subsidy.) The plan does not propose a revised success criterion or mission change, nor does it discuss whether the medical benefit justifies a subsidy model or whether the indication should pivot to spinal cord injury to reach viable scale.
1.5.B Tags
- target_miss_unaddressed
- commercial_viability_gap
- unit_economics_unstated
- mission_change_deferred
1.5.C Mitigation
By Month 12 (2027-10-31): commission an independent health-economic and eligible-population analysis covering: (a) ALS population aged 40–75 in EU/EEA, annual incidence, time-to-locked-in, eligible-pool size (current estimate: 500–1,400/year); (b) competing endpoints (communication vs. motor control) and willingness-to-pay from three HTA bodies (target: EUR 120,000–150,000/QALY); (c) spinal cord injury eligible pool (estimate: 2,000–5,000/year in EU) and motor-control outcome evidence; (d) centre capacity and procedural learning curve (10–20 implants/centre/year assumed, not validated). By Month 24 (2028-10-31), pre-register a 'Indication Pivot Gate': if eligible-ALS-pool analysis shows <800 per year and payers reject communication-only endpoints, mandate a pre-registered decision to switch to spinal cord injury or to accept a subsidy model. Document the subsidy model, if chosen: e.g., 'The device is a public-health good; the company is funded via grants and member-state co-payments, not private returns.' Define investor return expectations (IRR, horizon) up-front; if they require >10% IRR, the ALS programme is not viable and the company should not be formed.
1.5.D Consequence
The plan will run out of money or reach Month 180 (2041-10-31) with a company that has never been profitable and depends entirely on continued investor subsidy. This creates three risks: (1) Investors will lose confidence after Month 36–60 (2029-10-31 to 2031-10-31) when the commercial gap becomes visible, triggering a funding crisis that forces cost-cutting mid-programme (compromising safety or consent); (2) The company's continuity depends on multi-investor syndication, not a sustainable revenue model, which contradicts the 'patient-protection covenant' pledge of lifelong support; (3) If the company fails in Months 60–120 (2031-10-31 to 2036-10-31) when implanted patients exist, the patient-protection trust may be insufficiently sized to cover explantation, follow-up and replacement, violating the hard requirement of not abandoning patients.
1.5.E Root Cause
The plan treats 5,000 implants as a target but does not validate it against eligible-population epidemiology or payer willingness-to-pay. It reports the miss as a 'soft estimate' and moves on, without asking whether the indication or device positioning should change. The commercial-viability calculation (revenue vs. opex) is presented, but the implication—that the company is perpetually subsidy-dependent—is not addressed. The result is a plan that appears to proceed as if 1,050 is acceptable, yet still describes itself as having 'missed' 5,000, which is contradictory and masks a deeper decision: is this a commercial enterprise or a public-health programme?
1.6.A Issue - Leuven partner commitments are assumed, not secured; fallback sites are named but not pre-qualified; the plan lacks a go/no-go decision tree if primary partners decline.
The plan states 'imec, KU Leuven and UZ Leuven, none of which has committed to partner' and lists 'Ghent/Antwerp, Utrecht, Grenoble' as fallbacks without letters of intent, capacity assessments or cost estimates per site. The assessment says 'Secure hard commitments by Month 3 (2027-01-31)' but the plan does not specify what 'hard' means: a letter of intent? A resource-commitment memo? A signed MOU with specific deliverables (staff, facility hours, patient cohort) and decision timelines? If UZ Leuven (neurosurgery and ALS patient access) declines by Month 3 (2027-01-31), the plan says activate fallback by Month 6 (2027-04-30), with 'FIH +3–6 months', but does not say: which fallback is activated first? Are all three contacted in parallel (slowing decision) or sequentially (adding 3–6 months per decline)? What is the cost per fallback—imec has a EUR 300M cleanroom; Ghent/Antwerp hospitals have none. The plan also does not state what happens if a secondary commitment fails mid-way (e.g., imec's cleanroom capacity is consumed by existing projects by Month 9 = 2027-07-31).
1.6.B Tags
- partner_commitment_informal
- fallback_not_prequalified
- go_no_go_tree_missing
- resource_availability_unchecked
1.6.C Mitigation
By Month 1 (2026-11-30): send written partnership scopes to imec (cleanroom, electrode and ASIC prototyping, 12 months, estimated cost EUR 0.5–1M), KU Leuven (neuroengineering team secondment or contracting, 5 FTE, EUR 0.75–1.5M) and UZ Leuven (neurosurgery, operating theatres, ALS patient cohort, first-in-human feasibility, clinical trial hosting, estimated cost EUR 0.5–1M). Request signed LOIs by Month 3 (2027-01-31) specifying committed staff, facility hours (e.g., '200 hours/month imec cleanroom time'), patient-access commitments ('50 ALS patients screened by Month 24 = 2028-10-31'), and decision timeline ('board approval by 2027-01-15'). In parallel, contact Ghent/Antwerp, UMC Utrecht and CEA-Leti/Clinatec with identical scopes and request backup LOIs by Month 3 (2027-01-31). By Month 2 (2026-12-31), publish a 'Partner Decline Decision Tree': (a) if imec declines, activate CEA-Leti/Clinatec for cleanroom within 1 week (pre-identified contact, pre-quoted cost EUR 1–1.5M, 6-month lead time); (b) if UZ Leuven declines, activate UMC Utrecht within 1 week (pre-identified neurosurgery lead, pre-screened patient access); (c) if KU Leuven declines, contract external neuro-engineers within 2 weeks. At each activation, a 1-month clock starts; if the fallback LOI is not signed by Month 4 (2027-02-28), escalate to the board with a 'defer-or-stop' recommendation. Publish the decision tree to investors and the Commission before Month 1 (2026-11-30) ends.
1.6.D Consequence
Without pre-qualified fallbacks and a decision tree, the project enters a 'soft partner negotiation' state at Month 3 (2027-01-31) that could extend to Month 6 or 9 (2027-04-30, 2027-07-31), consuming founder time and delaying all technical and regulatory work. If a partner declines mid-programme (e.g., imec's cleanroom is over-allocated at Month 6 = 2027-04-30), the company faces a 3–6 month delay and a scramble to find alternative capacity, during which the FIH timeline slips, the team morale erodes, and early-stage investors become anxious about execution risk. The absence of cost pre-quotes per fallback means a secondary partnership could consume 20–30% of the T0 budget (EUR 15M), cascading into hiring freezes or missed gates.
1.6.E Root Cause
The plan lists the three Leuven institutions and three fallback sites as if partnerships are equivalent in cost and capability, but they are not. imec has industrial-scale cleanroom infrastructure; Ghent/Antwerp hospitals do not. UZ Leuven has an embedded ALS research clinic; other hospitals require patient-access agreements with ethics committees and delays. The plan does not pre-negotiate or pre-cost this variability, so partner selection appears flexible in the plan but is operationally rigid. The absence of a published decision tree means the founders will improvise responses if a partner declines, leading to delays and ad-hoc cost overruns.
2 Expert: Supply Chain & Manufacturing Sovereignty Lead (EU-Controlled Production, Pilot-Line Transfer, Dual-Sourcing)
Knowledge: Medical-device supply-chain strategy, pilot-line-to-validated-production transfer, dual-source qualification, export-control screening, EU/EEA supplier vetting, production-line validation and yield optimisation, quality systems transfer
Why: The plan rests on passing three EU-control tests on all components (ownership, maintainability without foreign consent, export-control freedom) by Month 72 (2032-10-31), with a critical path through design lock (Month 18 = 2028-04-30) → materials and process lock (Month 72 = 2032-10-31) → CE marking (Month 132 = 2037-10-31). The expert evaluates the Neural-Implant Foundry decision: whether the chosen dual-sourcing route for chips and substrates, in-house encapsulation and assembly, and pilot-line-to-production transfer are feasible, costed and scheduled correctly. They validate the upstream-dependency register (Month 12 = 2027-10-31) and advise on foundry-commitment fallbacks (Month 18 = 2028-04-30 gate).
What: Review the supply-chain architecture (which components are vertical, which dual-sourced, which licensed). Validate the sovereignty tests per supplier by Month 12 (2027-10-31): which foreign dependencies remain, and how are they mitigated? Confirm the pilot-line-to-Class-III transfer plan is budgeted in T1 (EUR 10M estimated) and gated before design lock. If the shared foundry (EUR 350M, 4+ member states) is not committed by Month 18 (2028-04-30), confirm the dual-sourcing fallback adds EUR 35M and 6-12 months, and cost it within T1/T2.
Skills: Supply-chain risk mapping and dual-source qualification, pilot-to-production transfer planning, ISO 13485 supplier audit and management, export-control compliance and restricted-items screening, manufacturing process lock and change control, yield analysis and contingency costing
Search: EU medical device supply chain sovereignty, dual sourcing qualification strategy, pilot line to production transfer, MDR supplier audit
2.1 Primary Actions
- Contact imec, KU Leuven and UZ Leuven by Month 0.5 (2026-11-15 14:00) with a detailed partnership scope (cleanroom access, neuroengineering teams, neurosurgery, ALS patient cohort, resource commitments, decision timelines); obtain written letters of intent from all three by Month 3 (2027-01-31 16:00), with backup LOIs from UMC Utrecht, Ghent/Antwerp and CEA-Leti/Clinatec also due Month 3 (2027-01-31). If any primary institution declines, activate the first-ranked fallback immediately (Month 3.5 = 2027-02-15) and compress site qualification to 12 weeks, with ethics pre-submission by Month 6 (2027-04-30 16:00).
- Identify and approach at least four candidate investor anchors by Month 1 (2026-11-30 16:00); present a 5-slide investment memorandum covering risk-retirement roadmap, return expectations (IRR, horizon, exit), capital per tranche and downside protection. Negotiate term sheets with at least two anchors by Month 3 (2027-01-31 16:00); secure written seed commitments (EUR 3–5M) by Month 3.5 (2027-02-15) and written T0 commitments (EUR 10–15M) by Month 9 (2027-07-31). If T0 is not committed by Month 11 (2027-09-30 16:00), pause non-critical hiring; if not closed by Month 12 (2027-10-31 16:00), defer FIH by 12 months and reduce team to 5 core.
- Commission an independent epidemiological analysis of eligible ALS and spinal-cord-injury populations across EU member states (annual prevalence, incidence, attrition, centre capacity) by Month 9 (2027-07-31). Initiate pre-reimbursement dialogues with HTA bodies in at least five countries by Month 12 (2027-10-31), presenting population data, health-economic endpoints and comparator technologies; request written payer willingness-to-pay thresholds by Month 18 (2028-04-30). If payers will not support EUR 120,000 per QALY, trigger a pre-registered indication pivot at the Month 24 (2028-10-31) feasibility gate.
- Establish an independent Data Monitoring Committee (DMC) by Month 6 (2027-04-30) with five or more members (two external statisticians, one patient representative, one ethics specialist, one external clinician); grant it binding pause authority. Pre-register quantitative safety, stability and bandwidth kill gates in writing by Month 3.5 (2027-02-15); publish them publicly before first-in-human. Tie investor tranches contractually to DMC written findings (e.g., Series B release conditional on DMC clearing chronic animal data by Month 24 = 2028-10-31).
- Recruit a consent advocate (independent clinical ethicist or patient representative) by Month 3 (2027-01-31); develop and pre-register a yes/no communication-validation protocol by Month 2.5 (2027-01-15), requiring two independent validators and video documentation. Submit the protocol to at least three ethics committees for pre-review by Month 3.5 (2027-02-15); obtain written approval before first-in-human. Establish quarterly re-consent checkpoints (baseline, monthly for 6 months post-implant, then quarterly). Any participant communication decline triggers immediate re-consent or exit offer.
2.2 Secondary Actions
- By Month 1 (2026-11-30 16:00), recruit or contract a CFO and a Chief Regulatory Officer. The CFO must own the financing timeline, gate verdicts and fallback activation; the CRO must own notified-body engagement (Month 6 = 2027-04-30), ISO 13485 QMS (Month 9 = 2027-07-31) and scientific-advice request (Month 12 = 2027-10-31). Issue a written hiring and facility roadmap by Month 2 (2026-12-31).
- Initiate a pre-submission meeting request with at least two notified bodies (TÜV Süd, BSI, DEKRA Assure) by Month 1 (2026-11-30 16:00), targeting meetings by Month 6 (2027-04-30). Present a Class III implantable neural-interface device, 18–24 month chronic animal study, first-in-human in Leuven and 1–2 other member states, decoder with AI Act high-risk classification and ISO 13485 QMS roadmap. Request written feedback on evidence gaps, resource availability and regulatory timeline by Month 7 (2027-05-31).
- Commission a freedom-to-operate (FTO) analysis for three candidate architectures (thin-film surface, penetrating cortical, endovascular) by Month 2 (2026-12-31); complete the analysis by Month 9 (2027-07-31). For each architecture, assess electrode patents (Neuralink, Synchron, universities), implant encapsulation IP, decoding algorithms and surgical-robotics IP. Screen all acquisition or licensing targets against the three EU-control tests (ownership, maintainability without foreign consent, export-control freedom) by Month 6 (2027-04-30). Ratify the architecture at Month 18 (2028-04-30) based on FTO, surgical risk, stability and bandwidth scoring.
- By Month 2 (2026-12-31), establish the ISO 13485 quality management system charter and assign a QA Manager. Initiate a Design History File (DHF) in a versioned, password-protected repository by Month 1 (2026-11-30 09:00). Implement a formal change-control process by Month 3 (2027-01-31): any design change requires a Change Request Form with risk assessment, validation plan and QA approval. All design work is logged in the DHF with traceability to design-review meetings or risk-management triggers.
- By Month 2 (2026-12-31), commission an AI Act compliance assessment from a biotech/medical-device AI legal specialist; deliver a written compliance roadmap by Month 9 (2027-07-31). By Month 18 (2028-04-30), define and lock the decoder version (Decoder Version 1.0 Pivotal Lock) for the first-in-human study. Pre-register an AI Act change-control process by Month 12 (2027-10-31): minor changes (retraining within locked feature set, <5% model weight change) do not require notified-body review; major changes (new features, new training data, architecture swap) require notified-body review. Budget EUR 2M for AI Act documentation and change-control administration over the programme.
- Establish exclusive field-limited licensing agreements with EU research institutes (imec, CEA-Leti, UMC Utrecht or equivalents) by Month 6 (2027-04-30), with pre-priced acquisition options converting no earlier than Month 24 (2028-10-31). Screen all licence targets for foreign consent rights, export-control exposure and third-party encumbrances. Sign supplier MOUs with at least two chip and two thin-film-substrate manufacturers by Month 12 (2027-10-31); qualify a second source for each by Month 30 (2029-04-30). Publish an upstream-dependency register (fab equipment, EDA software, test tools) by Month 12 (2027-10-31) with documented mitigation plans.
- By Month 3 (2027-01-31), commission a Leuven site resilience assessment covering multi-day heatwaves above 35°C, drought, flooding, power outages and cooling/water availability for cleanrooms, compute and wearable hardware. By Month 18 (2028-04-30), install backup power systems and redundant air-cooling; replicate critical compute at a second EU site by Month 24 (2028-10-31). Design the external processor to pass climate-chamber testing at 35–40°C with at least 12 hours battery life, passive cooling and automatic throttling that lowers bandwidth before skin or tissue temperature limits are approached.
- By Month 6 (2027-04-30), establish a patient-protection trust, funded separately from operating capital, with a EUR 5M floor before first-in-human and EUR 30,000 per implant at the time of implantation (total: EUR 150M if 5,000 implants; estimate EUR 31.5M for the central 1,050 cumulative). Contract a step-in agreement with an EU hospital network or second manufacturer, giving it escrowed designs, software source and spare-part rights if the company becomes insolvent. Commission an actuarial study of device lifetime and explantation costs by Month 9 (2027-07-31).
- By Month 12 (2027-10-31), issue a written competitive-response plan for the event that a US or Chinese device receives CE marking in the EU during Months 36–60 (2029-10-31 to 2031-10-31). The plan should define: (a) payer and hospital differentiation (EU data sovereignty, local service, lifelong support commitments); (b) pricing flexibility (options to reduce EUR 150,000 target if needed to compete); (c) a decision gate at Month 42 (2030-04-30) to assess market position and decide whether to continue, defer or pivot indication.
- By Month 15 (2028-01-31), complete pre-reimbursement dialogues with HTA bodies and develop health-economic models covering at least five member states. Build a health-economic case demonstrating cost per QALY at EUR 120,000 or below, comparing against eye-tracking, P300 and speech synthesis. Test the lead use case (home communication for people who have lost eye-tracking control) with patient organisations and payers by Month 18 (2028-04-30); measure their acceptance threshold. Develop outcome-based reimbursement contract templates by Month 24 (2028-10-31).
2.3 Follow Up Consultation
At Month 3 (2027-01-31) and Month 12 (2027-10-31) gates, review: (1) Leuven partnerships (written LOIs or fallback activation); (2) financing commitments (seed closes, T0 written, investor term sheets); (3) eligible-population epidemiology and payer willingness-to-pay (must be submitted before Month 18 = 2028-04-30); (4) FTO results and architecture selection (must be complete by Month 18 = 2028-04-30); (5) DMC charter, kill gates and consent protocol (must be published by Month 3.5 = 2027-02-15); (6) notified-body pre-submission feedback (expected Month 7 = 2027-05-31); (7) ISO 13485 QMS status and design-history file maturity. At Month 24 (2028-10-31), issue a final six-part feasibility verdict covering financing (committed versus hoped-for), population and reimbursement (payer validation), regulatory (notified-body alignment), supply-chain (foundry decision), clinical (consent readiness) and technical (architecture confidence). If financing, payer commitment or population viability is unresolved, defer implant programme and execute a research-only phase (Months 24–48 = 2028-10-31 to 2030-10-31) with reduced burn.
2.4.A Issue - Leuven partnership commitment is missing and poses an existential feasibility risk
The plan assumes imec, KU Leuven and UZ Leuven partnership by Month 3 (2027-01-31) but no written commitments exist at Month 0 (2026-10-31). This is not a secondary dependency; it is the foundation of the entire supply chain, clinical trial infrastructure and manufacturing roadmap. A decline by any institution at Month 3 (2027-01-31) would trigger a 3–6 month fallback activation and up to EUR 2M additional cost. The plan is operationally incoherent without these partnerships locked before MVP hiring and facility agreements are signed.
2.4.B Tags
- partnership_commitment_missing
- critical_path_risk
- month_3_deadline_unconfirmed
2.4.C Mitigation
By Month 0.5 (2026-11-15 14:00), contact imec, KU Leuven and UZ Leuven with a detailed partnership scope document specifying resource commitments, facility access, patient cohort access and decision timelines. Secure written letters of intent (not oral consent) by Month 3 (2027-01-31 16:00) from all three. In parallel, by Month 0.5 (2026-11-15), initiate equivalent pre-discussions with UMC Utrecht, Ghent/Antwerp University Hospital and CEA-Leti/Clinatec, requesting backup LOIs by Month 3 (2027-01-31 16:00). If any primary institution declines, activate the first-ranked fallback by Month 3.5 (2027-02-15) and compress site qualification to 12 weeks, with ethics pre-submission by Month 6 (2027-04-30 16:00).
2.4.D Consequence
Without written LOIs by Month 3 (2027-01-31), the programme cannot close seed funding (investors require confirmed partnerships), cannot sign facility leases or cleanroom access agreements, and cannot hire clinical and engineering staff with confidence. A delay in securing partnerships by 6 months compresses the entire first-in-human pathway by 6–12 months and forces design choices under schedule pressure. A partnership failure forces a 3–6 month fallback activation, moving FIH from Month 60 (2031-10-31) to Month 63–66 (2032-01-31 to 2032-04-30) and delaying CE marking by the same margin. This also undermines the financing narrative: investors will not commit T0 capital until partnerships are locked.
2.4.E Root Cause
Empty
2.5.A Issue - Financing gates lack written investor commitments and fallback mechanisms; the EUR 15M T0 and EUR 60M T1 are hoped-for, not secured
The plan defines financing gates at Months 12, 24 and 36 (2027-10-31, 2028-10-31, 2029-10-31) but states EUR 0 committed at Month 0 (2026-10-31). No investor return expectations, cost-of-capital assumptions or written pre-agreements exist. Gate G1 requires EUR 10–15M T0 in writing by Month 12 (2027-10-31), but there is no documented investor anchor, no pre-negotiated terms and no written fallback if the gate is missed. This violates the hard requirement that 'no funding commitments may be assumed'. The gate structure is aspirational, not feasible.
2.5.B Tags
- financing_gate_unfunded
- no_investor_anchor
- month_12_gate_at_risk
2.5.C Mitigation
By Month 1 (2026-11-30 16:00), identify and approach at least four candidate investor anchors (EU strategic investors, impact funds, or public financing institutions) with a 5-slide investment memorandum covering: (a) the risk-retirement roadmap and gate criteria; (b) investor return expectations (IRR, horizon, exit scenario); (c) capital requirements per tranche; (d) downside protection (patient-protection reserve, clear kill-gate authority). Negotiate term sheets with at least two anchors by Month 3 (2027-01-31 16:00), targeting EUR 3–5M seed closes by Month 3.5 (2027-02-15) and written T0 commitments by Month 9 (2027-07-31). Define fallbacks in writing: if T0 is not committed by Month 11 (2027-09-30 16:00), immediately pause non-critical hiring and defer full-scale animal studies to a 'reduced-scope' mode; if T0 is not closed by Month 12 (2027-10-31 16:00), automatically defer FIH by 12 months and reduce team to 5 core staff.
2.5.D Consequence
Without written investor commitments, the company cannot secure cleanroom or lab space, cannot hire the team documented in the plan (10 by Month 6 = 2027-04-30, 15 by Month 24 = 2028-10-31), and cannot fund chronic animal studies (budgeted at EUR 1.5–2M). Missing gate G1 at Month 12 (2027-10-31) forces a cascading deferral: FIH moves from Month 60 (2031-10-31) to Month 72–84 (2032-10-31 to 2033-10-31), CE marking slips from Month 132 (2037-10-31) to Month 144–156 (2038-10-31 to 2039-10-31) and the commercial case weakens further (fewer implants, lower cumulative revenue). A second missed gate (G3 at Month 24 = 2028-10-31) stops the implant programme entirely. The current gate structure is too permissive; it allows narrative renegotiation instead of automatic deferral or termination.
2.5.E Root Cause
Empty
2.6.A Issue - The 5,000-implant target is not credibly achievable and is already reported as missed; commercial viability and payer case are not grounded in eligible-population reality
The plan acknowledges that the 5,000-implant target is 'missed in all scenarios', with a central estimate of 1,050 cumulative implants by Month 180 (2041-10-31) and a break-even requiring 1,125 implants per year. However, the canonical eligible-pool estimate is 500–1,400 per year, implying a market size of only 7,500–21,000 people over 15 years across the entire EU/EEA. The ALS-only pool is at the lower end. Yet the reimbursement case, payer engagement strategy and investor return expectations are not documented against this constrained population. There is no demonstrated payer willingness to pay EUR 150,000 per implant or cost-effectiveness evidence at EUR 120,000 per QALY. This is the difference between an unproven clinical tool and a commercial medical device.
2.6.B Tags
- population_mismatch
- payer_case_undocumented
- reimbursement_assumptions_unvalidated
2.6.C Mitigation
By Month 9 (2027-07-31), commission an independent epidemiological analysis of ALS and high-spinal-cord-injury populations across EU member states, estimating annual eligible prevalence (age, severity, progression rate), annual incidence, attrition and centre/theatre capacity. Cross-check against published registries (e.g., EuroMND for ALS). By Month 12 (2027-10-31), initiate pre-reimbursement dialogues with HTA bodies in Germany, France, Italy, the Netherlands and Scandinavia, presenting: (a) the eligible-population epidemiology; (b) health-economic endpoints (QALYs, cost per QALY, cost per case); (c) comparator technology (eye-tracking, P300, speech synthesis) and measured daily-life benefit. Request written indication of payer willingness-to-pay thresholds and evidence gaps by Month 18 (2028-04-30). If payers will not commit to EUR 120,000 per QALY or higher, the cost-effectiveness case collapses and the indication strategy must change at the Month 24 (2028-10-31) feasibility gate.
2.6.D Consequence
The plan is incoherent if it pursues ALS-only without demonstrating payer acceptance and viable population scale. The current trajectory—1,050 implants, 300 implants per year at scale, an operating gap of EUR 66M per year—is unsustainable unless the company is treated as a public health infrastructure (subsidised indefinitely) or unless the indication expands to spinal cord injury or other conditions with larger eligible pools. Without documented payer commitment by Month 24 (2028-10-31), the feasibility verdict should be 'not credibly achievable' and the programme should either (a) defer commercial rollout and continue as a research/pilot programme, or (b) pivot to a broader indication (spinal cord injury, stroke, locked-in syndrome) that is decided at the pre-registered Month 78 (2033-04-30) gate. Proceeding to pivotal trials without this clarity risks abandoning patients if the company runs out of money at Years 12–15.
2.6.E Root Cause
Empty
3 Expert: Clinical Evidence & Health Economics (Reimbursement, Eligible Population, Payer Strategy)
Knowledge: Health technology assessment and payer negotiation, ALS epidemiology and natural history, assistive-technology market and reimbursement thresholds, cost-effectiveness and quality-adjusted-life-year modelling, clinical-endpoint design, indication-expansion strategy
Why: The plan's financial case hinges on whether EUR 150,000 price and 1,050 cumulative implants (estimated) by Month 180 (2041-10-31) support break-even (1,125 implants/year). The expert audits the First Indication and Outcome That Justifies Surgery decision (ALS communication vs. spinal cord injury), validates the eligible-population epidemiology (500-1,400 per year), and tests whether reimbursement and cost-effectiveness ceilings are payer-realistic. They advise on the pivotal indication gate (Month 78 = 2033-04-30), where the plan defers a commitment to ALS-only versus spinal cord injury. They also assess the communication-restoration outcome versus eye-tracking comparator credibility.
What: Commission or review an eligible-population epidemiology for ALS communication restoration across 3+ member states by Month 9 (2027-07-31). Initiate pre-reimbursement dialogues with payers in at least three member states by Month 15 (2028-01-31) to validate: (a) whether EUR 150,000 price is acceptable, (b) what daily-life benefit evidence is needed (beyond bandwidth), and (c) whether the indication should pivot to spinal cord injury. Model the revenue and break-even against the population forecast and implant ceiling.
Skills: HTA and reimbursement evidence requirements, ALS natural history and progression, comparative-effectiveness study design (communication technology, assistive devices), cost-effectiveness modelling and QALY estimation, payer negotiation and pricing strategy, eligible-population epidemiology and centre-capacity forecasting
Search: ALS patient population epidemiology Europe, brain computer interface reimbursement strategy, health technology assessment medical devices, assistive technology cost effectiveness
3.1 Primary Actions
- Contact imec, KU Leuven and UZ Leuven by Month 0.5 (2026-11-15) with a specific partnership proposal: imec provides cleanroom access and prototyping for electrodes and implant electronics; KU Leuven supplies neuroengineering and decoding teams; UZ Leuven provides neurosurgery, operating-theatre access and ALS patient cohort. Request written letters of intent by Month 3 (2027-01-31) confirming resource commitments, capacity (staff FTE, facility hours, patient referrals per year) and decision timelines. Simultaneously, contact three fallback sites (UMC Utrecht, Ghent/Antwerp, CEA-Leti/Clinatec) with the same scope and request preliminary capability letters by Month 3 (2027-01-31).
- Commission a formal health-economic model and eligible-population epidemiology by Month 9 (2027-07-31). Engage three member-state HTA bodies (e.g., equivalent to NICE, HAS, G-BA) by Month 15 (2028-01-31) with a draft reimbursement dossier. Test willingness to pay at EUR 150,000 and cost-effectiveness threshold. Validate eligible-population size (annual incidence, communication-impairment prevalence, surgical-centre capacity, patient consent likelihood) bottom-up by Member State. If payer feedback shows EUR 150k is unacceptable or eligible pool is <300 per year, adjust price or indication. Publish findings by Month 24 (2028-10-31) as a mandatory input to the pivotal-indication gate. Do not proceed to pivotal funding without payer acceptance in writing.
- Complete a 12-18 month chronic-implant animal study (minimum 6 animals, sheep or primate) by Month 18 (2028-04-30), documenting electrode signal decay, usable-channel count and thresholds for signal loss. Commission a literature synthesis of published thin-film, penetrating and endovascular signal-stability data from Clinatec, UMC Utrecht, Neuralink, Synchron and academic sources by Month 9 (2027-07-31). Convene a DMC working group to re-examine kill-gate thresholds (current: >20% animal loss blocks FIH, >25% human loss triggers pause) against this evidence and propose evidence-grounded alternatives by Month 15 (2028-01-31). Lock thresholds in the first-in-human protocol by Month 24 (2028-10-31). Publish a written revision of any threshold changes and the preclinical justification.
- Establish independent informed-consent safeguards for locked-in participants by Month 3 (2027-01-31). Recruit an independent consent advocate (clinical ethicist or patient-representative background) with binding authority to recommend participant exclusion if consent validity is questionable. Develop a yes/no communication-validation protocol with at least two independent validators and video documentation. Submit to ethics committees for pre-review by Month 3 (2027-01-31); obtain written approval by Month 6 (2027-04-30). Establish a consent re-confirmation schedule (baseline, monthly for 6 months, quarterly for 12 months, then annually) and offer re-consent or exit options if communication declines.
- Assign a clinical-liaison lead by Month 1 (2026-11-30) responsible for parallel engagement of primary Leuven sites and three fallback sites. Define a fallback-activation protocol: if a primary site declines by Month 3 (2027-01-31), activate the matched fallback by Month 4 (2027-02-28) and compress site qualification to 12 weeks (ethics pre-submission by Month 6 = 2027-04-30). Cost fallback activation at EUR 0.5-2M (estimate). If more than one primary site declines, trigger the financing gate G1 early (Month 12 = 2027-10-31) and publish a revised feasibility verdict.
- Secure EUR 5M in seed funding and national grants by Month 3 (2027-01-31) and T0 committed in writing (floor EUR 10M) by Month 12 (2027-10-31). Define three hard gates with written fallbacks: G1 (Month 12 = 2027-10-31) T0 EUR 10-15M committed, backed by architecture choice and notified-body pre-submission; G3 (Month 24 = 2028-10-31) EUR 60M of T1 committed and feasibility verdict published; G4 (Month 36 = 2029-10-31) EUR 210M of T1+T2 committed or FIH deferred in 12-month steps. Publish a written verdict (proceed, defer, or switch) within 5 business days of each gate. Failure to meet gate criteria triggers automatic deferral, not renegotiation.
3.2 Secondary Actions
- Engage a notified body by Month 6 (2027-04-30) and request a pre-submission meeting (target Month 4-5 = 2027-02-28 to 2027-03-31) specifying Class III implantable neural interface, first-in-human at Month 60 (2031-10-31), CE marking at Month 132 (2037-10-31), and decoder as AI Act high-risk. Request written feedback on evidence gaps and regulatory timeline by Month 6 (2027-04-30). Request Commission scientific advice on clinical investigation design and AI Act scope by Month 12 (2027-10-31).
- Establish an independent Data Monitoring Committee by Month 6 (2027-04-30) with at least five members (two statisticians, one patient representative, one ethics specialist, one external clinician). Grant binding authority to recommend pause or stop. Define and publish quantitative kill gates before implantation: SAE threshold ≥2 in first 10 or ≥3 in first 20 implants; signal loss >25% in 12 months; sustained bandwidth <5 bits/s in ≥50% of participants. Contractually bind investor tranches to DMC findings.
- Build an ISO 13485 quality-management system and design-history file by Month 9 (2027-07-31). Assign a Quality Assurance Manager by Month 2 (2026-12-31) with MDR experience. Initiate the DHF by Month 1 (2026-11-30) before any design work. Implement change control, traceability and document versioning. Request a preliminary QMS audit from the notified body by Month 18 (2028-04-30); incorporate findings by Month 24 (2028-10-31).
- Define AI Act high-risk compliance and decoder change-control thresholds by Month 9 (2027-07-31). Commission an external AI Act assessment covering training-data provenance, model-validation protocols, bias-testing plans and post-CE change-control scope. Lock a decoder version and training-data boundary for the pivotal trial by Month 18 (2028-04-30); document as Decoder Version 1.0 (Pivotal Lock) in the DHF. Pre-register change-control categories (minor, major) and notified-body review triggers. Allocate EUR 0.5-1M budget for AI Act compliance over the programme.
- Complete freedom-to-operate analysis per architecture candidate (thin-film surface, penetrating cortical, endovascular) by Month 9 (2027-07-31). Commission an independent FTO review and use results as a scoring input for architecture ratification at Month 18 (2028-04-30). Screen licence targets against the three EU-control tests (ownership, maintainability without foreign consent, export-control freedom) before pricing acquisition options. Publish an upstream-tool dependency register (fab equipment, EDA software, IP blockers) by Month 12 (2027-10-31).
- Secure written partnership commitments from two EU chip and substrate suppliers by Month 12 (2027-10-31). Define a dual-sourcing plan: each supplier independently qualifies to Class III standards and supply-contract terms include price-cap, multi-year volumes and escrow of design files. Qualify the second source by Month 30 (2029-04-30). If a shared foundry is not committed by Month 18 (2028-04-30), immediately switch to the dual-sourcing route and fund transfer-to-production at approximately EUR 10M (T1) + EUR 25M (T2).
- Assess Leuven site climate resilience by Month 3 (2027-01-31): heatwaves above 35°C, drought, flooding, power outages. Install backup power and redundant cooling by Month 18 (2028-04-30). Replicate EU-hosted compute at a second site. Design the external processor for at least 12-hour battery life with passive cooling and automatic throttling (tested at 35-40°C). Screen trial participants for impaired thermoregulation; issue heat-alert protocols.
- Recruit a founding team of 10 by Month 6 (2027-04-30): neurosurgeon, two neuroengineers, clinical-trials manager, regulatory affairs specialist, QA manager, supply-chain engineer, finance manager, two additional senior researchers. Offer equity (5-10% founder pool) and multi-year salary with 3-5 year retention bonuses (minimum EUR 20k each) to offset 20-30% EU pay deficit. Cross-train deputies by Month 10 (2027-08-31). Establish weekly all-hands meetings and quarterly in-person retreats.
3.3 Follow Up Consultation
Next consultation should focus on three topics: (1) Payer and HTA outcome: report results of engagement with three member-state HTA bodies by Month 18 (2028-04-30). If payers accept EUR 150k and cost-effectiveness ceiling, lock it into the pivotal plan; if not, present alternative prices or indications and re-cost the commercial case. (2) Chronic-stability preclinical data and revised kill-gate thresholds: present animal-study results, literature synthesis and DMC-recommended threshold changes by Month 18 (2028-04-30), and lock them in the first-in-human protocol by Month 24 (2028-10-31). (3) Leuven partner status: report written LOIs or fallback activations by Month 6 (2027-04-30). At Month 24 (2028-10-31), publish a final six-part feasibility verdict: scientific, engineering, clinical, regulatory, supply-chain and financing. If financing remains unconfirmed, deferral to a narrower ALS-feasibility phase (Months 24-48 = 2028-10-31 to 2030-10-31) should be named explicitly.
3.4.A Issue - Reimbursement and Payer Case Completely Unvalidated; Market-Access Strategy Is Speculative
The plan proposes EUR 150,000 per implant and EUR 120,000 per QALY but reports zero payer or HTA engagement to date. No health-economic model exists; cost per QALY is unmodeled and unvetted. The commercial case (EUR 45M revenue vs. EUR 111M opex at Year 15, EUR 66M annual gap) assumes payer reimbursement without evidence that three or more member states will accept either the price or the indication. Eligible-population epidemiology is missing; the 1,050-implant estimate lacks bottom-up validation against centre capacity, reimbursement coverage and actual patient demand.
3.4.B Tags
- missing-payer-validation
- unmodeled-health-economics
- speculative-commercial-case
- no-epidemiology-anchor
- financial-break-even-unrealistic
3.4.C Mitigation
Commission a formal health-economic model and eligible-population epidemiology by Month 9 (2027-07-31). Engage three member-state HTA bodies (e.g., UK NICE equivalents, Haute Autorité de Santé, G-BA or equivalent) by Month 15 (2028-01-31) with a draft reimbursement dossier. Test willingness to pay at EUR 150,000 and cost-effectiveness threshold tolerance. Bottom-up eligible-population validation: count annual ALS incidence (1-2 per 100k across EU), estimate communication-impairment prevalence (15-20% of ALS), and screen against centre surgical capacity and patient consent likelihood. Adjust price or indication if payer feedback shows EUR 150k is unacceptable or if eligible pool is <300 per year. Publish findings by Month 24 (2028-10-31) as a mandatory input to the pivotal-indication gate.
3.4.D Consequence
Without payer validation, the commercial case remains speculative. The company risks designing and manufacturing for a market that rejects the price or lacks sufficient eligible patients. If payers demand EUR 50-80k instead, the EUR 111M annual opex cannot be covered by 1,050 implants and the company will run out of reserves. Alternatively, the indicated population may be 50% smaller than estimated, making 1,050 impossible and forcing closure before or shortly after CE marking. This is a leading cause of medtech failure: clinical success without reimbursement.
3.4.E Root Cause
The plan was built around the long-term bandwidth mission and sovereignty requirements, not around reimbursement feasibility. Payer engagement and health-economic modeling were deferred as secondary, even though they determine whether the 5,000-patient (or revised 1,050-patient) business case is viable. The plan notes the gap but does not gate capital release on resolving it.
3.5.A Issue - Chronic Signal Stability Kill Gates Lack Preclinical Grounding; Threshold Credibility Uncertain
The plan proposes that >20% signal loss in animals blocks FIH and >25% in humans triggers pause, but provides no published animal or human data supporting these thresholds. No chronic rat, primate or human evidence is cited. The thin-film surface architecture (provisional until Month 18 = 2028-04-30) has limited long-term published data outside Clinatec (WIMAGINE) and UMC Utrecht. If preclinical work shows 15-20% loss is normal for thin-film, the gate threshold becomes meaningless; if 30%+ loss is observed, architecture switch costs EUR 30M and 12-24 months, compressing the entire timeline.
3.5.B Tags
- unjustified-kill-thresholds
- no-preclinical-validation
- architecture-risk-underspecified
- thin-film-precedent-weak
- gateway-credibility-absent
3.5.C Mitigation
Before Month 18 (2028-04-30), complete a 12-18 month chronic-implant animal study (minimum 6 implants, sheep or primate) documenting electrode signal amplitude, noise floor, usable-channel count and their decay over 12 months. Commission a literature synthesis of published thin-film, penetrating and endovascular signal-stability data from Clinatec, UMC Utrecht, Neuralink, Synchron and academic sources by Month 9 (2027-07-31). Convene a Data Monitoring Committee working group to re-examine kill thresholds against this evidence and propose evidence-grounded alternatives by Month 15 (2028-01-31). If preclinical data shows thresholds must be relaxed (e.g., 30% loss is normal), update gates and publish the revision; if data supports the threshold, lock it in the first-in-human protocol by Month 24 (2028-10-31). Plan for a contingency cost of EUR 30M if architecture switch is triggered.
3.5.D Consequence
Kill gates without grounded thresholds lose regulatory and ethics credibility. A notified body or ethics committee may reject a gate that looks arbitrary. Alternatively, if thresholds are too tight, the company risks being forced into an unplanned EUR 30M architecture switch at Month 24-36 (2028-10-31 to 2029-10-31), compressing FIH and delaying CE marking by 12-24 months. If thresholds are too loose, the programme may continue with unstable electrodes that fail clinically, leading to patient harm and regulatory shutdown. Either way, the decision loses authority and the programme loses time or credibility.
3.5.E Root Cause
The plan sets kill gates without having generated the preclinical data needed to justify them. It assumes that a 20-25% threshold is reasonable, but this is an engineering guess, not an evidence-grounded target. The thin-film surface architecture was chosen before chronic-stability data existed. The plan needs preclinical data to lock the threshold, not vice versa.
3.6.A Issue - No Fallback if Leuven Partners Decline; Clinical-Site Risk Unmanaged Until Month 6 (2027-04-30)
The plan states that imec, KU Leuven and UZ Leuven 'have not committed' but lists fallback sites (Ghent/Antwerp, Utrecht, Grenoble) as activated 'by Month 6 (2027-04-30) if any declines.' However, there is no written protocol for deciding which fallback activates, no cost or delay estimate per declined partner, and no pre-qualification of the fallback site's surgical capacity or patient cohort. If UZ Leuven (the only confirmed ALS implanting centre in the plan) declines, Utrecht's neurosurgery must be qualified in 12 weeks—a compressed timeline with unspecified ethics, training and consent-protocol setup. The dual-track approach (parallel LOI outreach) is mentioned but not resourced or scheduled before Month 3 (2027-01-31).
3.6.B Tags
- partner-commitment-uncertain
- fallback-site-unqualified
- clinical-site-risk-unmanaged
- timeline-compression-unspecified
- contingency-cost-underestimated
3.6.C Mitigation
By Month 1 (2026-11-30), assign a clinical-liaison lead responsible for engaging all three primary sites (imec, KU Leuven, UZ Leuven) and the three fallback sites (Ghent/Antwerp, Utrecht, Grenoble) in parallel. By Month 3 (2027-01-31), obtain written letters of intent from at least two of the three primary sites specifying resource commitments (cleanroom hours, surgeon FTE, patient-referral pathway) and decision deadlines. Simultaneously, request preliminary capability letters from each fallback site (ALS patient census, surgeon experience, ethics-committee contact, timeline to readiness). Define a fallback-activation protocol: if a primary site declines by Month 3 (2027-01-31), activate the matched fallback by Month 4 (2027-02-28) and compress site qualification to 12 weeks (ethics pre-submission by Month 6 = 2027-04-30, surgeon training by Month 9 = 2027-07-31). Cost the fallback activation at EUR 0.5-2M (estimate) depending on distance, regulatory re-submissions and surgery-team setup. If more than one primary site declines, trigger the financing gate G1 (Month 12 = 2027-10-31) early and publish a revised feasibility verdict.
3.6.D Consequence
Without a managed fallback, a decline from UZ Leuven forces a crisis-mode scramble to qualified another centre, loses 3-6 months in the FIH timeline (compressing the entire downstream schedule), and incurs unknown costs for site qualification and ethics approval. The company may miss the Month 60 (2031-10-31) FIH window, pushing CE marking from Month 132 (2037-10-31) to Month 144-156 (2038-10-31 to 2039-10-31). Alternatively, the company may activate a fallback that lacks ALS referrals or experienced surgeons, compromising the quality of the first cohort and creating safety or ethical risk. Either outcome damages the financing case at gate G1 (Month 12 = 2027-10-31) or G3 (Month 24 = 2028-10-31).
3.6.E Root Cause
The plan treats Leuven partnerships as the default and fallback sites as contingencies, without acknowledging that 'no commitment' means a significant risk. The founders have not yet approached any site, so the confidence that imec, KU Leuven and UZ Leuven will partner is untested. The fallback sites are named but not pre-qualified or integrated into the early-stage work. A founder-only team is unlikely to have capacity to manage four parallel site-engagement tracks in Month 0-3 (2026-10-31 to 2027-01-31) without additional clinical leadership.
4 Expert: Patient Safety & Ethics Governance (Informed Consent, Data Monitoring, Clinical Trial Oversight)
Knowledge: Clinical-trial ethics and informed-consent protocols for vulnerable populations, independent data-monitoring committees, serious-adverse-event thresholds and stopping rules, GDPR and neural-data governance, longstanding implant follow-up and patient registries, bioethics in emerging neurotechnology
Why: The plan designates consent and monitoring as hard requirements with binding authority granted to a Data Monitoring Committee (DMC) by Month 6 (2027-04-30). The expert validates the Living Consent decision (independent advocate, re-confirmation schedule, yes/no signalling validation) and pre-registers kill gates (2 or 3 serious adverse events, 25% signal loss, below 5 bits/s in 50%, invalid consent at a site). They ensure the DMC charter, member selection, decision rules and publication timelines are credible and independent. They also advise on consent language when communication is impaired and how to reconcile early enrolment (while speech is intact) with the greatest-need principle.
What: Draft and review the informed-consent template and yes/no validation protocol by Month 2 (2026-12-31); submit to ethics committees for pre-review by Month 3 (2027-01-31). Recruit or contract the independent consent advocate by Month 3 (2027-01-31). Charter the DMC (5+ members, patient representative, patient-advocate organisations) by Month 6 (2027-04-30), with published decision rules and a monthly safety-review process. Pre-register quantitative kill gates (SAE thresholds, signal stability, bandwidth floor, consent validity) with the board and investors before Month 3 (2027-01-31), and ensure each gate is tied to a financing tranche (e.g., Series B release conditional on DMC safety clearance).
Skills: Informed-consent design for cognitively impaired and communication-restricted populations, independent data-monitoring committee governance and charter, serious-adverse-event classification and causality assessment, clinical-trial stopping rules and decision analytics, GDPR medical-data handling and EU hosting requirements, patient-representative and advocacy-group engagement
Search: informed consent locked-in syndrome ALS, data monitoring committee medical device trials, clinical trial ethics vulnerable populations, serious adverse event stopping rules
4.1 Primary Actions
- By Month 2 (2026-12-31), commission an independent consent-research study with ethics partners (advocate, lead ethics committees in Belgium, Netherlands, France) to develop a capacity-fluctuation monitoring tool and a binding advance-directive form for ALS participants, with video-recording protocols and clinical-team enforcement rules. Secure written ethics-committee pre-approval of the consent protocol by Month 4 (2027-02-28) before participant recruitment; budget EUR 100k.
- By Month 4 (2027-02-28), establish a written Data-Monitoring Committee charter with formal AI Governance Subcommittee authority to audit decoder versions, training-data lineage and model-validation compliance monthly. Define patient-representative selection via ALS-advocacy group, binding escalation to notified body and board within 7 and 3 days respectively on AI Act breach detection, and contractual tranche conditioning on quarterly DMC clearance. Budget EUR 250k per year for independent AI audit staff.
- By Month 3 (2027-01-31), commission a EUR 150k actuary-and-governance study defining the independent trust structure (custodian, insolvency-remote jurisdiction, IP-collateral backup), revised per-implant reserve (likely EUR 40,000–60,000), and funding timeline tied to trial enrollment. Lock reserve-governance terms in Series A by Month 12 (2027-10-31); make each commercial-implant sale deposit its reserve amount before revenue recognition.
- By Month 3 (2027-01-31), secure written confirmation from imec, KU Leuven and UZ Leuven on specific consent-safeguard roles (consent advocate oversight, re-confirmation scheduling, advance-directive handling). If imec or KU Leuven cannot provide ethics-research infrastructure for the consent study, engage fallback partners (UMC Utrecht, Ghent University, CEA-Leti) by Month 2 (2026-12-31).
- By Month 6 (2027-04-30), submit the finalised consent protocol and DMC charter to the notified body as part of pre-submission engagement; request written feedback on consent-validity criteria and decoder-audit expectations. If the notified body flags unresolved consent or AI governance gaps, resolve them before design lock (Month 18 = 2028-04-30).
4.2 Secondary Actions
- Develop a participant-communication handbook (for patients and families) explaining the advance-directive form, the re-confirmation schedule, the independent consent advocate's role, and the conditions under which the study can be paused or the device switched off, in plain language and video format. Pilot-test this with ALS patient groups by Month 9 (2027-07-31).
- Establish a quarterly 'Patient Safety Review' forum open to implanted patients, families, the consent advocate and the DMC, where safety concerns and device changes are discussed transparently. Publish redacted summaries to all stakeholders within 30 days. This builds trust and gives implanted patients a standing voice beyond the formal DMC.
- Commission an independent ethics review of the reserve-governance structure by a bioethics centre (e.g., Cambridge Bioethics, Erasmus Institute of Bioethics) to test whether the proposed trust design satisfies the ethical and legal requirements of patient protection. Complete by Month 9 (2027-07-31).
- Define clear escalation pathways for the consent advocate and DMC to raise concerns directly to ethics committees, the notified body and the Commission without company interference. Publish these pathways in writing by Month 6 (2027-04-30); include a 'whistleblower' clause protecting the advocate and DMC members from retaliation.
- Conduct a mock AI Act compliance audit (by external auditor, not company staff) of the planned decoder change-control system by Month 15 (2028-01-31), before the first decoder version is locked. Use findings to refine the AI Governance Subcommittee process and budgeting.
- Model the reserve adequacy under stress scenarios: company bankruptcy, prolonged device failures, explantation wave, and care-escalation (e.g., participants becoming non-verbal and needing full-time caregiver coordination). Publish a 'stress-test report' by Month 12 (2027-10-31) showing reserve coverage against each scenario.
4.3 Follow Up Consultation
"At the next consultation (Month 6 = 2027-04-30), report: (1) consent protocol finalisation, ethics-committee pre-approvals and any modifications required; (2) DMC charter signed, patient representative selected, and AI Governance Subcommittee staffing and budget confirmed; (3) trust-governance and actuary study completed, revised per-implant reserve amount justified, and Series A terms drafted; (4) notified-body feedback on consent and AI audit expectations, and any design implications; (5) imec/KU Leuven/UZ Leuven written statements on consent-safeguard roles and resource commitments, or fallback-site activation in progress. By Month 12 (2027-10-31), present the feasibility verdict on consent and governance: can the plan enrol locked-in ALS participants with binding consent validity and unambiguous device-control wishes, or does the population's communication loss require narrowing the indication (e.g., earlier-stage ALS only, or a different indication)? If consent feasibility is uncertain, the Month 24 (2028-10-31) scope gate must address it explicitly."
4.4.A Issue - Informed consent safeguards are underspecified for rapid disease progression and communication loss in ALS
The plan mandates an independent consent advocate and re-confirmation checkpoints (baseline, monthly for 6 months, then quarterly), but does not address three critical consent challenges in ALS: (1) participants may lose communication capacity mid-trial before a re-consent window arrives; (2) capacity fluctuation (good days and bad days) may undermine the validity of a single yes/no validation; (3) no binding mechanism exists to convert pre-declared participant wishes (e.g., 'switch off if signal drops below 5 bits/s') into enforceable protocol deviations. The plan acknowledges consent complexity but defers specifics to Month 3–5 (2027-01-31 to 2027-03-31) ethics-committee review, which risks late discovery of consent-method misalignment during first-in-human.
4.4.B Tags
- informed-consent-gaps
- vulnerable-population
- ALS-communication-loss
- advance-directives
- ethics-committee-alignment
4.4.C Mitigation
By Month 2 (2026-12-31), co-develop a dynamic consent protocol with the independent consent advocate and ethics partners: (1) adopt a capacity-fluctuation monitoring tool (e.g., day-of-week communication reliability, fatigue tracking) that flags days when re-consent is unreliable, and document when consent cannot be validly obtained; (2) implement a binding pre-implant advance-directive form, witnessed and video-recorded, that specifies participant-chosen device-off, explantation or study-exit triggers tied to measurable outcomes (bandwidth, signal stability, adverse events) with clear instructions for the clinical team; (3) secure written ethics-committee pre-approval of the advance-directive format and enforcement pathway by Month 4 (2027-02-28), before participant recruitment. Add EUR 100k (estimate) for consent-research infrastructure and monthly re-validation audits.
4.4.D Consequence
Without specificity, the plan risks enrolling participants whose wishes are not validly recorded or actionable, leading to ethics-committee challenges, trial pauses (proposed threshold: any invalid-consent finding pauses a site), and reputational damage if participants feel unheard or coerced. Advanced ALS participants are acutely sensitive to autonomy; absent clear advance directives, the company may face legal liability if a participant's death or deterioration occurs after unclear consent. This could delay or halt the trial, consuming months of the Month 12 (2027-10-31) financing gate.
4.4.E Root Cause
The plan treats consent as a governance checkpoint rather than a lived, participant-controlled process. ALS participants in advanced stages experience rapid, unpredictable communication loss; the fixed re-confirmation schedule cannot adapt to this volatility. Pre-declared conditions are mentioned but not operationalised, so the plan risks implanting participants whose wishes cannot be reliably elicited or honoured post-implant.
4.5.A Issue - Data-Monitoring Committee governance lacks participant representation and binding escalation triggers for AI Act violations
The plan establishes an independent DMC by Month 6 (2027-04-30) with 5+ members including one patient representative, holding binding pause authority on safety. However, the plan does not specify: (1) how the patient representative is selected (organisation affiliation, conflict-of-interest criteria); (2) what happens if the DMC detects an AI Act violation (e.g., decoder retraining using non-consented data, feature drift, model staleness) before a formal change-control gate is triggered; (3) how often the DMC reviews decoder version changes and training-data lineage; (4) what escalation path exists if the company ignores a DMC safety recommendation. The plan's proposed safety thresholds (≥2 SAEs in first 10, ≥3 in first 20 implants) are event-based, not process-based, so silent consent or data-handling failures may not trigger a review until harm occurs.
4.5.B Tags
- data-monitoring-governance
- patient-representation-quality
- AI-Act-enforcement-gap
- decoder-change-control
- regulatory-escalation
4.5.C Mitigation
By Month 4 (2027-02-28), establish a written DMC charter specifying: (1) patient representative selection via an ALS-advocacy group (e.g., European ALS Federation) with a signed independence declaration and quarterly reselection; (2) a standing 'AI Governance Subcommittee' (two DMC members plus company's chief data officer) that reviews all decoder version changes, training-data sources and model-validation reports monthly, with independent verification that changes comply with the locked change-control thresholds (e.g., 'minor' retraining within locked feature set does not require notified-body review); (3) a binding escalation: if the subcommittee finds an AI Act breach (unauthorised data use, undocumented model swap, training on out-of-scope participants), it must notify the notified body within 7 days and the board within 3 days, with public reporting within 30 days; (4) investor tranches are contractually conditioned on quarterly DMC audits finding no substantive violations. Fund this at EUR 250k per year (estimate) for independent AI audit staff.
4.5.D Consequence
Without AI Act audit integration into DMC oversight, the plan risks deploying high-risk decoders that violate transparency, auditability and change-control obligations, triggering notified-body enforcement action during pivotal dossier review or post-CE-marking. Such a finding could force device recall or market withdrawal, destroying patient trust and clinical credibility. An under-resourced patient representative may also fail to raise safety concerns, eroding the ethical legitimacy of the consent process and ethics-committee support. This could delay CE marking by 12–24 months and expose the company to regulatory penalties.
4.5.E Root Cause
The plan underestimates the DMC's role in continuous AI governance. The decoder is high-risk under the AI Act, and the plan envisions a 'locked change-control process' (Decoder Stack Residency), but the DMC has no standing authority to audit decoder lineage, training data, or model updates. Patient representation is mandated but not operationalised, so the voice of implanted patients may be token rather than binding. Escalation to regulators or the board is assumed but not contractually defined.
4.6.A Issue - Patient-protection reserve mechanism is specified but not funded, gated or audited to ensure survival through device lifetime
The plan mandates a EUR 30,000 per-implant reserve in a trust outside investor reach, set up by Month 6 (2027-04-30) with a EUR 5M floor. However, the plan does not specify: (1) who administers the trust (independent trustee, hospital, third party), what governance protects it from investor seizure or insolvency law disputes, or what triggers release of funds (annual audit of company balance sheet, device failure event, company bankruptcy filing); (2) how the reserve is actuarially sized (expected device lifetime is 10-15 years, estimate, but no actuarial studies of ALS survival post-implant, explantation risk or service-cost trajectory exist); (3) what happens if the company fails before accumulating the full reserve (e.g., only 150 trial implants = EUR 4.5M against the EUR 5M floor, so the reserve is under-funded); (4) how the reserve is invested (cash holds value, equities expose it to market risk) and whether real-estate or IP is escrowed as collateral. The plan calls it a 'hard requirement' but treats it as a secondary financial detail.
4.6.B Tags
- patient-continuity-obligation
- reserve-governance
- insolvency-protection
- actuarial-uncertainty
- escrow-administration
4.6.C Mitigation
By Month 3 (2027-01-31), commission a governance-and-actuary study (EUR 150k, estimate) covering: (1) establishment of an independent trust structure under EU law (multistate jurisdiction or EU-level fiduciary) with a custodian (e.g., a major EU pension fund, hospital endowment or dedicated medical-device trust) that is legally insolvency-remote and holds the power to release funds only on documented need (annual audit showing company depletion below a cash threshold, bankruptcy filing, device failure event); (2) actuarial modelling of ALS patient survival post-implant (using existing ALS databases), expected device lifetime (10-15 years), explantation rates, service costs (follow-up visits, replacements, storage) and contingencies, yielding a revised per-implant reserve (likely EUR 40,000–60,000, estimate); (3) a collateral-backup mechanism: the company grants the trustee a lien on IP (decoder, encapsulation design, manufacturing know-how) that can be liquidated or transferred to a successor manufacturer if the company becomes insolvent, ensuring continuity without needing cash. Lock this in the Series A term sheet by Month 12 (2027-10-31). Allocate total trust funding at Month 0 (2026-10-31): EUR 5M (floor for 150 trial implants) plus a commitment that each commercial implant contributes its reserve amount before sales revenue is recognised.
4.6.D Consequence
Without binding governance and actuarial grounding, the reserve is a promissory note, not a guarantee. If the company faces financial stress at Month 100–120 (2035-02-28 to 2036-10-31) (2035–2036), when the first cohort is aging and requiring explantation, the company may pressure the board to treat the reserve as operating capital, leaving later patients without support. Implanted patients and their families may then face legal battles over who is responsible, reputationally damaging the device and undermining public trust in EU medical-device governance. Regulators may also condition CE marking on proof that the reserve is truly protected, so a vague structure discovered during dossier review could force rework or denial.
4.6.E Root Cause
The plan assumes the reserve can be set up by Month 6 (2027-04-30) but does not address the governance and legal risk that make it credible. EU insolvency law varies by member state; a trust set up in Belgium may not be ring-fenced against creditor claims in a French or Dutch bankruptcy. Actuarial sizing is guessed (EUR 30,000 per implant) without justification. The plan also does not address the political pressure that an insolvent company may face to raid the reserve to fund final clinical operations, which would leave later patients unprotected.
5 Expert: Founder and Chief Executive Officer (Strategy, Governance, Stakeholder Alignment)
Knowledge: Deep-tech venture founding and scaling, medical-device company governance, investor relations and capital raising, European institutional affairs and Commission engagement, strategic decision-making under uncertainty and kill-gate authority
Why: The plan is unfunded at Month 0 (2026-10-31) and requires a founder-led team to secure written commitments at five gates (G1 Month 12 = 2027-10-31, foundry Month 18 = 2028-04-30, G3 Month 24 = 2028-10-31, G4 Month 36 = 2029-10-31, pivotal Month 78 = 2033-04-30). The CEO/founder validates the governance model: whether stop, defer or switch authority is credibly independent, how the DMC and consent advocate are protected from pressure, and whether the board structure and investor terms allow gates to be called honestly. They also assess whether the founding team can sustain three years (Months 0-36 = 2026-10-31 to 2029-10-31) of hiring and site-negotiation against a 20-30% EU pay deficit while managing partner refusal, financing failure, architecture pivots and public scepticism.
What: Validate the governance charter by Month 1 (2026-11-30): who holds binding stop authority (board, DMC, investors), what happens if they disagree, and how gate verdicts are published. Establish the founding team recruitment and retention plan by Month 2 (2026-12-31), with named fallbacks if key hires decline. Secure written LOIs from imec, KU Leuven and UZ Leuven by Month 3 (2027-01-31), with parallel fallback-site talks. Close at least EUR 5M seed by Month 3 (2027-01-31) and T0 (EUR 10M floor) by Month 12 (2027-10-31), with gate verdicts published within 5 business days. Map stakeholder influence (Commission, member states, notified bodies, payers, public) and set engagement milestones at Months 6, 12, 24 and 36 (2027-04-30, 2027-10-31, 2028-10-31, 2029-10-31).
Skills: Venture capital and institutional fundraising, company formation and cap table, board governance and fiduciary duty, European Commission and member-state liaison, stakeholder mapping and influence strategy, decision authority under uncertainty, media and public engagement, talent recruitment and retention against international pay differentials
Search: deep tech venture CEO medical devices, European Commission engagement IPCEI, investor relations capital raising healthcare, stakeholder governance European startups
5.1 Primary Actions
- By Month 2 (2026-12-31), engage a CFO or capital-raising partner with Series A and B medical-device experience, and commission binding term-sheet templates for T0 (EUR 10-15M seed) and T1 (EUR 60-120M) that tie tranches to documented gate criteria (architecture lock, chronic animal data completion, notified-body engagement confirmation) with explicit loss-of-control and renegotiation rules. Begin investor outreach to 3-5 anchor investors (institutional, impact or corporate) by Month 1 (2026-11-30) with LOI targets by Month 3 (2027-01-31). Document investor board composition (seats per tranche, patient representative, independent scientist) and define the CEO removal and gate-override processes in the board charter.
- By Month 9 (2027-07-31), commission an independent health-economics and epidemiology study for three member states covering ALS population (incident, prevalent, eligible, mortality, centre capacity), high cervical spinal cord injury (incidence, centre capacity, surgical complexity) and stroke motor recovery (expansion option). Model break-even volume and implant-centre capacity per indication at 5, 10 and 25 sites. Hold pre-reimbursement dialogues with payers in France, Germany and the Netherlands (statutory insurance, HTA bodies) by Month 15 (2028-01-31) to establish evidence requirements and price tolerance. Publish a revised commercial case by Month 18 (2028-04-30) with three scenarios (ALS-only, SCI-primary, stroke-future) and explicit T1 and T2 implications for each.
- By Month 1 (2026-11-30), hold face-to-face meetings with imec, KU Leuven and UZ Leuven specifying partnership scope (cleanroom access, neuroengineering secondment, theatre time and patient cohort, resource commitments, IP terms, decision gate Month 6 = 2027-04-30). In parallel, conduct fallback negotiations with UMC Utrecht, Ghent/Antwerp neurosurgery, and CEA-Leti/Clinatec; obtain preliminary willingness by Month 2 (2026-12-31) and fallback LOIs by Month 6 (2027-04-30). If a primary site declines by Month 3 (2027-01-31), immediately activate the preferred fallback site and compress qualification to 12 weeks (site visit, ethics pre-submission, surgeon credentialing by Month 6 = 2027-04-30). Document site-equivalence criteria (cleanroom Class, theatre capacity, ALS cohort size, insurance indemnity) and fallback cost (EUR 1-1.5M, 6-12 weeks delay) before Month 3 (2027-01-31).
5.2 Secondary Actions
- By Month 3 (2027-01-31), recruit a founding team of at least 10 (lead neurosurgeon, two neuroengineers, clinical-trials manager, regulatory specialist, quality-assurance manager, supply-chain engineer, finance manager and two cross-trained deputies). Offer equity packages (5-10% founder pool) and multi-year salary commitments with EUR 20k+ retention bonuses at 3-year and 5-year marks to offset the 20-30% EU pay gap versus US competitors. Establish a weekly all-hands meeting and quarterly in-person retreats to sustain cohesion across distributed sites.
- By Month 2 (2026-12-31), hire or contract a Quality Assurance Manager with ISO 13485 and MDR experience. Establish a quality-management system (QMS) charter covering document control, design-history file (DHF), training, supplier audit and change control. Initiate the DHF in a secure, versioned repository before any design work. By Month 6 (2027-04-30), implement formal change-control procedures for all design changes, with documented risk assessment and QMS traceability.
- By Month 6 (2027-04-30), establish an independent Data Monitoring Committee (5+ members: two external statisticians, one patient representative, one ethics specialist, one external clinician) with binding authority to pause, modify or recommend stop. Pre-register quantitative kill-gate criteria before any implantation (SAE threshold ≥2 in first 10 or ≥3 in first 20; signal stability >25% loss in 12 months; sustained bandwidth <5 bits/s in ≥50% of participants). Bind investor tranches contractually to DMC written findings (e.g., Series B release conditional on DMC safety clearance).
- By Month 1 (2026-11-30), recruit an independent consent advocate (clinical ethicist or patient-representative background) to attend every consent conversation and validation, hold binding authority to recommend exclusion, and remain independent of the company. By Month 2 (2026-12-31), develop and pre-register a yes/no communication-validation protocol with independent validators and video recording. Submit the consent protocol to ethics committees (Leuven, Utrecht, Grenoble) for pre-review by Month 3 (2027-01-31); incorporate feedback and obtain written approvals by Month 5 (2027-03-31).
- By Month 3 (2027-01-31), close at least EUR 5M in seed funding and secure written LOIs from T0 investors (floor EUR 10M, target EUR 15M) for Month 12 (2027-10-31) close. Define the three hard financing gates: G1 Month 12 (2027-10-31) with EUR 3-5M Series A conditional on architecture choice, notified-body engagement and ≥50% chronic animal data; G3 Month 24 (2028-10-31) with EUR 60M T1 committed and feasibility verdict published; G4 Month 36 (2029-10-31) with EUR 150M T2 committed or FIH deferred. At each gate, publish a written verdict (proceed, defer, switch) within 5 business days; missed gates trigger automatic deferral, not renegotiation.
- By Month 6 (2027-04-30), submit pre-submission meeting requests to at least two notified bodies (e.g., TÜV Süd, BSI, DEKRA Assure) for Class III neural implant with first-in-human Month 60 (2031-10-31) and AI Act decoder risk. Schedule meetings for Month 4-5 (2027-02-28 to 2027-03-31). Establish ISO 13485 QMS by Month 9 (2027-07-31). Request Commission scientific advice by Month 12 (2027-10-31) referencing notified-body feedback. Conduct preliminary QMS audit by Month 18 (2028-04-30) before design lock.
- By Month 9 (2027-07-31), commission an AI Act compliance assessment from an external legal specialist (biotech or medtech AI). Deliver a written compliance roadmap by Month 9 (2027-07-31). Establish locked decoder version and training-data boundary for the pivotal trial by Month 18 (2028-04-30). Pre-register AI Act change-control thresholds by Month 12 (2027-10-31) (minor changes require no notified-body review; major changes require scientific review and possible clinical re-validation). Allocate EUR 0.5-1M budget for AI Act documentation and change-control administration.
- By Month 3 (2027-01-31), establish the patient-protection reserve trust with a EUR 5M floor before first-in-human, funded separately from operating capital and outside investor reach. Commit EUR 30,000 per implant, 100% funded at implantation. Set up quarterly trust administration and rebalancing. Estimate device lifetime and explant-and-support obligations by Month 12 (2027-10-31) using actuarial methods.
5.3 Follow Up Consultation
At Month 12 (2027-10-31) and Month 24 (2028-10-31), the board must present: (1) written tranches signed or with signed LOIs (T0 by Month 12 = 2027-10-31, T1 at least EUR 60M binding commitment by Month 24 = 2028-10-31, or automatic deferral); (2) epidemiology and payer-acceptance results with revised commercial case and indication pivot scenario; (3) confirmed Leuven partnerships or activated fallback sites with signed MOUs for theatre time, patient access and resource commitments; (4) DMC charter with published kill-gate criteria and first safety results; (5) notified-body written feedback on clinical investigation design and AI Act scope; (6) finalized architecture choice with FTO analysis, chronic-animal data ≥70% complete, and bilateral supplier MOUs with second-source qualification plan. If any of these five dependencies is unresolved by Month 24 (2028-10-31), the decision rule is automatic deferral to Month 36 (2029-10-31) focus on narrow ALS communication feasibility (Months 24-48 = 2028-10-31 to 2030-10-31, EUR 30-50M budget), with a binary go/no-go decision at Month 48 (2030-10-31) on whether to proceed to pivotal-trial funding. This protects the company from the trap of continuing to burn capital on an unfunded, unfocused, unconfirmed multi-year programme.
5.4.A Issue - Financing structure is unmanned and governance of capital gates is absent—investors will walk if tranches are not contractually binding with explicit loss-of-control consequences
The plan assumes EUR 0 committed at Month 0 (2026-10-31), then expects T0 (EUR 10-15M) by Month 12 (2027-10-31) without naming a single investor, governance structure, or return term. Gates G1, G3, G4 are paper exercises: 'published verdicts' within 5 days mean nothing if a funder disagrees. Investors will demand board seats and veto rights; the plan forbids assuming funding yet forbids reducing evidence. Either the plan accepts dictatorial founder control (unworkable with capital partners), or it must pre-negotiate binding term sheets with anchor investors by Month 3 (2027-01-31), including investor composition of the board and explicit loss-of-control terms if gates are missed. Without investor signatures, not words, capital gates collapse.
5.4.B Tags
- governance-vacuum
- investor-misalignment
- capital-stage-mismatch
- gate-credibility
5.4.C Mitigation
By Month 2 (2026-12-31), hire a CFO or capital partner (not a contractor) with Series A and B medical-device raising experience. Commission a term-sheet template for T0 (EUR 10-15M seed) and T1 (EUR 60-120M) that explicitly binds the tranches to written gate criteria (architecture lock, chronic animal data, notified-body engagement). Name 3-5 anchor investors (institutional, impact or corporate) and initiate outreach by Month 1 (2026-11-30) with a target of signed LOIs by Month 3 (2027-01-31). Define investor board composition (e.g., 1-2 investor seats for T0, 3-4 for T1 plus a patient representative and independent scientist) and loss-of-control rules if a gate fails. If investors are not locked by Month 6 (2027-04-30), pivot to entirely grant-funded MVP (reduce scope to chronic animal studies and architectural validation only) and defer Series A to Month 12 (2027-10-31). Document the governance trades explicitly in the board charter; a sceptical investor will ask 'who fires the CEO if the Q2 2028 (Month 24 = 2028-10-31) gate is missed?' and expect a written answer.
5.4.D Consequence
If governance is not resolved by Month 2 (2026-12-31), Series A will slip to Month 18-24 (2028-04-30 to 2028-10-31) or later because investors will demand clarity on control and loss-of-control consequences. Every month of delay compresses the architecture-validation and chronic-animal schedules. If capital gates are not contractually binding, founders will attempt to continue past a failed gate (momentum inertia), investors will demand renegotiation, and the programme will oscillate between stop and defer for months, burning 10-20% of the tranche in overhead. Patient-protection reserves become secondary in this chaos: the question 'do we have money for Year 2?' will dominate board discussion. The plan's credibility with regulators and ethics committees also collapses; they will see a programme with unclear financing and ask 'will you still be here if trials go poorly?'
5.4.E Root Cause
The plan conflates 'hard requirements' (patient safety, EU control) with 'soft targets' (implant numbers, bandwidth) but treats financing as a soft assumption that can be 'hoped-for'. In reality, capital is a hard requirement: without committed tranches, no hiring, no cleanroom access, no animal studies. The gate structure is written as if the founders own 100% and can enforce stops; real investors will demand governance power tied to their tranches, and if founders cannot deliver a gate, investors will demand board negotiation, not a published 'deferral verdict'. The contradiction is unresolved.
5.5.A Issue - The 5,000-implant and EUR 150,000 price-per-implant reimbursement case is internally inconsistent—eligible population epidemiology is missing, payer acceptance is untested, and the plan does not explain how a 300-implant-per-year ceiling in Year 15 scales back to a break-even of 1,125 implants per year
The plan projects about 1,050 cumulative implants and 550 active users by Month 180 (2041-10-31), explicitly admitting the 5,000 target is 'missed in all scenarios'. Yet it proposes EUR 150,000 per-implant pricing and EUR 120,000 per QALY cost-effectiveness, without showing who will pay it, at what volume, in which member states. The implied Months 168-180 (2040-10-31 to 2041-10-31) capacity is 300 implants per year across about 25 centres (12 per centre), but no eligible-population epidemiology validates this. If ALS-only, about 500-1,400 eligible per year; for high cervical spinal cord injury, about 200-300 per year. Either the population is too small (ALS alone cannot sustain 300/year), or the plan must pivot to a broader indication (e.g., stroke motor recovery, which is 50,000+ per year in the EU, but requires a different surgical target and consent narrative). The plan gates this decision at Month 78 (2033-04-30), late enough that T2 and T3 tranches are committed.
5.5.B Tags
- market-size-mismatch
- payer-case-untested
- population-epidemiology-missing
- indication-pivot-timing
5.5.C Mitigation
By Month 9 (2027-07-31), commission an independent health-economics and epidemiology study for three member states (France, Netherlands, Germany) covering: (1) ALS population size (incident, prevalent, eligible), annual mortality, disease trajectory and centre capacity for implantation; (2) high cervical spinal cord injury population, annual incidence, surgical centre capacity and rehabilitation length of stay; (3) stroke motor recovery (later expansion option): incidence and centre willingness to implant for motor control. Model implant capacity at 5, 10, and 25 centres per indication, and estimate break-even volume by indication and pricing. By Month 15 (2028-01-31), hold pre-reimbursement dialogues with payers in France, Germany and the Netherlands (statutory health insurance bodies, HTA authorities) to establish: (a) which indication(s) they would fund; (b) evidence requirements (RCT sample size, comparator, primary endpoint); (c) price tolerance (cost per QALY). Publish a revised commercial case by Month 18 (2028-04-30) with three scenarios (ALS-only, SCI-primary, stroke-future), each with explicit implications for T1 manufacturing investment and T2 centre rollout. Make the Month 78 (2033-04-30) indication gate a go/no-go on commercial-case evidence, not design or safety evidence.
5.5.D Consequence
If epidemiology and payer acceptance are not resolved by Month 18 (2028-04-30), Series A investors will demand it before committing to T1 (EUR 120M), slipping T1 to Month 30 (2029-04-30) at the earliest. If the plan admits that 5,000 is unrealistic but does not revise the fundraising narrative, Series B (EUR 480M) will face investor scepticism about unit economics and founder credibility. Payers will also wait until late-stage trial data to commit; if the commercial case is unclear at Month 24 (2028-10-31) (Month 78 = 2033-04-30 decision gate is far away), the pivotal trial will start without payer agreement, and reimbursement negotiation will begin only after CE marking, delaying revenue by 12-24 months. The plan's profitability gap (EUR 66M per year at Month 180 = 2041-10-31) is then a company failure, not a soft miss. If ALS alone is the viable indication, the EUR 2.5B envelope is oversized, the team is larger than needed, and the programme should be resized to EUR 400-800M and repositioned as a 'niche communication platform', not a 'mass-market EU BCI'.
5.5.E Root Cause
The plan was written to satisfy a user constraint (5,000 implants) and a planning ceiling (EUR 2.5B), but the feasibility verdict then admits the target is unrealistic. Rather than revise the target or acknowledge that the scope must change, the plan treats the reimbursement and indication as secondary, gated decisions, leaving them unresolved. Investors and payers will not fund a EUR 2.5B programme on the assumption of 'decided at Month 78 (2033-04-30)' if that decision determines whether the venture is profitable. The implicit commercial logic is: if ALS alone is too small, then spinal cord injury is the only viable indication, but the plan provides no evidence (FTO, reimbursement, surgeon adoption) that spinal cord injury is credible. The plan is thus either a EUR 2.5B bet on an indication that will be decided late (unacceptable to capital partners), or it is a smaller EUR 500M-1B programme focused on ALS communication (and it should say so).
5.6.A Issue - Leuven partnerships are not confirmed, fallback sites are not pre-qualified, and the plan does not explain what happens operationally if imec or UZ Leuven declines after Month 3 (2027-01-31)—the 'EUR 2M cost, 3-6 months delay' estimate is unverified and insufficient to guide a fallback decision
The plan states 'none of which has committed to partner' but requires letters of intent from imec, KU Leuven and UZ Leuven by Month 3 (2027-01-31). It names fallbacks (Ghent/Antwerp, Utrecht, Grenoble) but provides no evidence that they are equivalent in cleanroom capacity, neurosurgery expertise, ALS patient cohorts or willingness to partner. If UZ Leuven declines, the plan suggests activating Utrecht or Ghent by Month 6 (2027-04-30), but does not specify: Who leads site selection? Is it a competitive pitch or a pre-negotiated fallback? What are the non-negotiable requirements (cleanroom Class 100 or Class 1000? implant-capable OR theatre only? ALS cohort 50+ or 10+)? If the fallback site has different surgical protocols or patient demographics, do the ethics approvals, consent templates and trial design change? The cost estimate ('about EUR 2M') is vague: facility setup, legal, training or all three?
5.6.B Tags
- partnership-contingency-untested
- operational-fallback-unclear
- site-equivalence-unverified
- delay-estimate-not-grounded
5.6.C Mitigation
By Month 1 (2026-11-30), hold a face-to-face meeting with imec (cleanroom director and business development), KU Leuven (neuroengineering department chair) and UZ Leuven (neurosurgery chief and hospital executive responsible for clinical trials). Present the partnership scope: imec = cleanroom access for electrode microfabrication and prototype assembly, hands-on training, and intellectual-property acknowledgement; KU Leuven = seconded postdoctoral engineer (2 FTE) for 3 years for decoding and neural-signal analysis; UZ Leuven = theatre time (1 day per month, ramping to 2-3 days per month by Month 84 = 2033-10-31), ALS patient-cohort access, neurosurgery lead (0.2-0.5 FTE), ethics liaison, and insurance indemnity. Request written LOIs by Month 3 (2027-01-31) with specific resource commitments and a decision gate (Month 6 = 2027-04-30, allow 3 months for institutional approvals). Simultaneously, conduct parallel fallback negotiations with: UMC Utrecht (contact neurosurgery chief and hospital R&D officer), Ghent University Hospital and Antwerp University Hospital (contact joint neurosurgery team), and CEA-Leti/Clinatec (contact clinical director). Ask each for a preliminary willingness statement by Month 2 (2026-12-31) and a full fallback LOI by Month 6 (2027-04-30). Distinguish first-in-human requirements (one site, 10-20 implants) from pivotal multi-site requirements; only the first-in-human site must be locked by Month 6 (2027-04-30). If a primary site declines by Month 3 (2027-01-31), immediately escalate to the preferred fallback (e.g., UZ Leuven declines → activate UMC Utrecht at Month 3.5 = 2027-02-15) and compress the fallback-site qualification to 12 weeks (site visit, ethics pre-submission, surgeon credentialing plan by Month 6 = 2027-04-30). Document the cost and delay: facility setup (cleanroom fees, lab build-out) EUR 0.3-0.5M, legal and insurance EUR 0.2-0.3M, surgeon training EUR 0.2-0.4M, regulatory re-submission EUR 0.1-0.3M, total 'about EUR 1-1.5M' (not EUR 2M); schedule slip is 6-12 weeks, not 3-6 months, if the fallback is pre-qualified.
5.6.D Consequence
If partnerships are not confirmed by Month 3 (2027-01-31) and fallbacks are not pre-qualified, a decline by a primary partner will create 2-3 months of urgent site negotiation and approval cycles (Month 3-6 = 2027-01-31 to 2027-04-30 crisis), during which the team is paralysed and other workstreams (chronic animal studies, supplier engagement) stall. If the fallback is not pre-qualified, the new site will demand ethics approvals and regulatory pre-submission by Month 6 (2027-04-30), which is feasible only if drafted and submitted by Month 5 (2027-03-31), leaving no time for feedback iteration. The first-in-human start will slip from Month 60 (2031-10-31) to Month 66-72 (2032-04-30 to 2032-10-31), which compresses the timeline from pivotal lock (Month 72 = 2032-10-31) to CE marking (Month 132 = 2037-10-31) by 6-12 months, forcing design shortcuts and faster manufacturing transfer that increase risk. Patient recruitment also suffers: ALS is progressive; every month delays include more advanced, shorter-lived patients, reducing post-implant follow-up duration and limiting safety data.
5.6.E Root Cause
The plan treats partnerships as binary (committed or not) but does not map the operational dependencies. Leuven is chosen for proximity to the Commission and to imec/KU Leuven, but the plan forbids assuming any of these as committed. If UZ Leuven declines, does the plan move to Utrecht (Netherlands, multicentre coordination) or Ghent (Belgium, single-country regulatory path)? Each choice has different implications for ethics approvals, surgeon training and trial sequencing. The plan also does not distinguish between 'host hospital for first-in-human' (one site, months 60-84 = 2031-10-31 to 2033-10-31) and 'pivotal-trial sites' (5-15 sites, months 84-132 = 2033-10-31 to 2037-10-31). A fallback for first-in-human is urgent; fallbacks for pivotal sites can be arranged later. By combining them, the plan creates urgency where it may not exist.
6 Expert: Bandwidth Measurement and Competitive Benchmarking Lead (Bits-per-Second Protocol, Data Residency, Rival Tracking)
Knowledge: Neural decoding and information-transfer-rate measurement, held-out task protocols and independent verification, federated learning and EU data residency, competitive intelligence on US and Chinese BCI systems, machine-learning model validation and reproducibility
Why: The plan rests on measuring net user-controlled information transfer rate (bits/s), independent verification inside the EU (data never leaves), and annual scenarios for rival best-systems (25/60/150 bits/s pessimistic/central/optimistic, estimates). The expert validates the Bits-per-Second Claim Discipline decision: whether the pre-registered protocol with held-out tasks, independent replication and separate reporting of command selection versus free communication are feasible and will survive scrutiny from regulators, payers and a sceptical public. They also advise on whether the target of at least 10 bits/s in at least 3 participants is realistic for thin-film surface architecture and whether the bandwidth kill floor (below 5 bits/s in 50% of participants, proposed threshold) should trigger a switch to penetrating cortical.
What: Commission or conduct a literature review of published thin-film surface and penetrating cortical bandwidth claims by Month 9 (2027-07-31), including Clinatec WIMAGINE, UMC Utrecht, Neuralink, Synchron and any Chinese publications, and build a factual baseline for the rival 25/60/150 scenario. Pre-register the information-transfer-rate protocol (task suite, measurement window, error correction, user-control verification) by Month 15 (2028-01-31). Design or contract an independent verification laboratory (EU-based) and federated-test infrastructure where decoder code is sent to the data, and only aggregate metrics leave the EU site. Establish a monthly rival-tracking update process, and re-baseline the pessimistic/central/optimistic scenarios every 12 months from public disclosures.
Skills: Information-theory and Shannon entropy measurement, neural decoding algorithm validation, held-out test design and statistical power, federated learning and secure multi-party computation, competitive intelligence gathering and synthesis, EU data-residency infrastructure and governance, machine-learning model reproducibility and open science
Search: information transfer rate BCI measurement protocol, brain computer interface bandwidth benchmarking, neural decoding validation independent verification, federated learning medical data EU
6.1 Primary Actions
- By Month 6 (2027-04-30), establish a Bandwidth Measurement and Verification Working Group under a written charter; pre-register the information-transfer-rate protocol with an external registry (Open Science Framework or equivalent) by Month 9 (2027-07-31); allocate €500k–€1M for a secure federated testing platform by Month 12 (2027-10-31); and integrate the protocol into preclinical animal studies so kill-gate thresholds are your own data, not external guesses.
- By Month 9 (2027-07-31), commission an independent systematic review of thin-film and penetrating-microelectrode chronic-implant signal-retention data; revise the 20% animal kill threshold based on published evidence; assign a Deputy CTO to lead the penetrating-cortical contingency design in parallel by Month 6 (2027-04-30); and schedule a formal Data Interpretation meeting by Month 18 (2028-04-30) with the DMC present.
- By Month 24 (2028-10-31), have a validated health-economic model covering eligible ALS population, daily-life functional outcome, QALY impact and cost-effectiveness; conduct pre-reimbursement dialogues with 3+ member-state HTA bodies by Month 15 (2028-01-31); establish a Payer Advisory Board by Month 12 (2027-10-31); and pre-register a hard Indication Pivot Decision at Month 24 (2028-10-31) tying continuation to population and reimbursement evidence (do not defer to Month 78 = 2033-04-30).
6.2 Secondary Actions
- Commission a published competitive-intelligence report on best US and Chinese BCI systems by Month 15 (2028-01-31), sourcing from peer-reviewed literature, FDA/NMPA submissions and public demonstrations; establish update cycles tied to rival announcements so the rival scenarios (25/60/150 bits/s) are refreshed quarterly.
- Consult with information theorists and signal-processing specialists from EU labs (e.g. imec, TU Delft, University of Tübingen) on the held-out task battery and error-correction methods; ensure the bandwidth protocol is aligned with Shannon entropy and published neural-decoding benchmarks to withstand peer review and regulator scrutiny.
- Engage the notified body on bandwidth (information transfer rate) as an essential performance criterion in the MDR dossier by Month 12 (2027-10-31) pre-submission; confirm whether the kill thresholds (5 bits/s floor, 10 bits/s target) are acceptable or whether the notified body has tighter requirements.
- Build a contingency cost model for the penetrating-cortical fallback: update the EUR 30M, 12–24 month estimate with actual design, tooling and validation work by Month 18 (2028-04-30); clarify whether the fallback architecture uses electrode arrays matching your existing bench tests or requires new material qualification.
- Request that the Payer Advisory Board (to be established) define the daily-life outcome measure jointly with clinical and patient partners; run a small feasibility study or focus-group testing with ALS patients and caregivers by Month 18 (2028-04-30) to validate that the proposed outcome (e.g. 4+ hours/day home communication independence) is meaningful to them.
- Develop a sensitivity analysis showing how changes in eligible-population size (±50%), reimbursement price (±25%), centre throughput (±20%) and device lifetime (±5 years) affect cumulative implants and company breakeven; publish ranges by Month 24 (2028-10-31) rather than point estimates, so investors and regulators understand the true uncertainty.
6.3 Follow Up Consultation
"At Month 15 (2028-01-31), present the Bandwidth Measurement and Verification Working Group's locked protocol, the systematic review of signal-retention data, the revised kill thresholds, the health-economic model, and early payer feedback. At that point, confirm whether federated testing infrastructure is ready for Month 18–24 (2028-04-30 to 2028-10-31) preclinical work, whether the chronic animal studies are on track to inform the architecture decision, and whether the Indication Pivot Decision at Month 24 (2028-10-31) can be credibly made. If any pillar (measurement, stability, reimbursement) is not ready, escalate the Month 24 (2028-10-31) decision to Month 36 (2029-10-31) and defer non-critical spending under the gate mechanism. At Month 24 (2028-10-31), expect to report: (a) locked bandwidth protocol with independent replication plan, (b) chronic animal data with interpreted kill-gate outcome, (c) validation or revision of signal-stability kill thresholds, (d) health-economic model accepted or challenged by payers, (e) eligible-population epidemiology confirmed or shown to be unviable, and (f) written go/no-go decision on ALS-first indication or pivot to spinal cord injury. The credibility of the feasibility verdict depends on all six."
6.4.A Issue - Bits-per-Second Measurement Protocol Lacks Independent Verification Infrastructure
The plan targets ≥10 bits/s net (user constraint) and claims independent verification (Decision 15), but specifies no secure compute environment, no federated testing consortium, no held-out task protocol, and no multi-centre replication framework. The bandwidth kill floor (5 bits/s in 50% of participants, proposed threshold) is set without preclinical evidence supporting its realism. The rival scenarios (25/60/150 bits/s, estimate) are unsourced and unvalidated. Without locked measurement protocols and EU-hosted verification infrastructure in place by Month 18 (2028-04-30), claims will be dismissed as cherry-picked or biased, undermining regulator and payer trust precisely when CE dossier credibility is critical.
6.4.B Tags
- bandwidth-verification-gap
- federated-testing-absent
- rival-benchmarking-unsourced
- held-out-protocol-undefined
6.4.C Mitigation
By Month 6 (2027-04-30), establish a Bandwidth Measurement and Verification Working Group (5–7 members: information theorists, independent EU labs, patient representatives, one regulator observer) under a written charter that locks the information-transfer-rate protocol, the held-out task battery, error-correction methods and independent replication rules. Pre-register the protocol with an external registry (e.g., Open Science Framework) by Month 9 (2027-07-31). Allocate €500k–€1M (budget line) for a secure federated testing platform (data stays in the EU, code travels) by Month 12 (2027-10-31). Commission a published competitive-intelligence report on best US and Chinese systems by Month 15 (2028-01-31), sourcing from peer-reviewed literature, FDA submissions and public demos, with update cycles tied to rival announcements. Integrate the locked protocol into the preclinical animal study (Months 12–24 = 2027-10-31 to 2028-10-31) so the kill-gate thresholds are grounded in your own data, not external guesses. Engage the notified body on bandwidth as part of the essential performance criteria (MDR Annex I) by Month 12 (2027-10-31) pre-submission.
6.4.D Consequence
At Month 60 (2031-10-31) FIH, early bandwidth claims will lack independent replication, regulator confidence and public credibility. Payers and ethics committees will demand re-verification, delaying CE dossier assembly by 6–12 months (estimate). Rival CE-marked devices entering the EU market at Months 36–60 (2029-10-31 to 2031-10-31) will claim superior bits/s with unverified headlines; without a published, locked protocol and independent verification results, this company's claims will be perceived as defensive or retrospective. Investors will lose confidence in the bandwidth narrative, which underpins the long-term mission and the payer case.
6.4.E Root Cause
The plan treats bandwidth as a secondary technical outcome ('Bits-per-Second Claim Discipline', Decision 15, Medium priority) rather than as a foundational measurement framework integral to the regulatory dossier and competitive positioning. No budget is assigned to building verification infrastructure; no timeline assigns Month 18 (2028-04-30) architecture lock as the deadline for locking the measurement protocol. Decision 15's three options (EU-only verification, federated consortium, rival scenarios) are presented as alternatives rather than mandatory together.
6.5.A Issue - Chronic Signal Stability Kill Gate Lacks Preclinical Evidence Foundation
The plan proposes kill thresholds (20% animal signal loss blocks FIH; 25% human loss in 12 months pauses; 50% retention below 36 months triggers switch, proposed thresholds) without citing any preclinical rodent or non-human primate chronic-implant data showing that thin-film surface electrodes or the fallback penetrating microelectrodes will meet these targets. Published thin-film surface data (e.g., WIMAGINE at Clinatec, UMC Utrecht's ALS home-use study) shows variable 6–12 month signal retention (50–90%, estimate, not verified); no source confirms that 80% retention at 12 months is achievable in your architecture. The plan budgets 18–24 months for chronic animal studies (Months 0–24 = 2026-10-31 to 2028-10-31) but assigns no decision date for interpreting results, no pre-registered success criteria, and no contingency if results show >20% loss in month 6–12 (2027-04-30 to 2027-10-31).
6.5.B Tags
- stability-threshold-unsourced
- animal-study-timeline-vague
- signal-loss-kill-criteria-arbitrary
- preclinical-interpretation-gate-missing
6.5.C Mitigation
Commission an independent systematic review of published thin-film and penetrating-microelectrode chronic-implant data (rodent, non-human primate, human) by Month 3 (2027-01-31), focusing on signal retention, impedance rise and unit isolation over 6–36 months. Publish it by Month 9 (2027-07-31) with a table of success rates and failure modes per architecture candidate. Revise the kill thresholds by Month 9 (2027-07-31) based on this review; if published data shows thin-film surface achieving only 70% 12-month retention (estimate), set the animal gate at 65% (to allow a small margin), and commit in writing that penetrating-cortical work begins by Month 20 (2028-06-30) if animal results breach this by Month 18 (2028-04-30). Schedule a formal Data Interpretation and Go/No-Go meeting by Month 18 (2028-04-30), with the DMC (Decision 14) and an independent statistician present, to decide architecture fate within 5 business days of final animal data. Assign a Deputy CTO to lead the penetrating-cortical contingency design by Month 6 (2027-04-30) in parallel, so fallback work is not a surprise.
6.5.D Consequence
If chronic animal studies show 25–30% signal loss within 12 months (a realistic risk for thin-film surface electrodes, estimate), the team will face a Month 18–24 (2028-04-30 to 2028-10-31) binary choice: proceed despite failing the 20% kill gate (eroding the credibility of pre-registered criteria), or switch to the penetrating-cortical fallback at an estimated cost of EUR 30M and 12–24 months delay, compressing architecture validation into Months 24–48 (2028-10-31 to 2030-10-31) and pushing FIH to Month 72–90 (2032-10-31 to 2034-10-31). The pivotal trial and CE marking will slip to Month 156+ (2039-10-31+), colliding with competitive US/Chinese CE marks and funding tranches (T2/T3, Months 48–84 = 2030-10-31 to 2033-10-31) already drawn.
6.5.E Root Cause
The plan adopts worst-case safety thresholds (20% blocks FIH) without validating whether the chosen architecture can meet them. The sourcing route (exclusive licences with options on unknown IP) means the team has not yet benchmarked the predecessor devices' signal stability. Chronic-animal studies are scheduled but not tied to specific go/no-go decision dates; results will likely arrive Month 18–24 (2028-04-30 to 2028-10-31) alongside architecture lock, creating a collision between data interpretation and irrevocable design commitments.
6.6.A Issue - Eligible-Population Epidemiology and Reimbursement Credibility Absent; 5,000-Implant Target Not Validated
The plan acknowledges that the 5,000-implant target is missed in all scenarios and reports 1,050 central implants (estimate) as likely by Month 180 (2041-10-31). However, no bottom-up epidemiology of eligible ALS patients, no centre/theatre capacity model, no payer willingness-to-pay study and no reimbursement thresholds are documented. The proposed price (EUR 150,000 per implant) and cost-effectiveness ceiling (EUR 120,000 per QALY) are thresholds only, with no payer endorsement. The eligible pool estimate (500–1,400 per year, estimate) is derived from no cited source. Without these foundations, the financing case (revenue EUR 45M per year, opex EUR 111M per year, operating gap EUR 66M per year at Month 180 = 2041-10-31, estimates) is speculative, and the decision to proceed with an ALS-first indication (Decision 3) rests on no validated clinical or commercial evidence.
6.6.B Tags
- epidemiology-unvalidated
- payer-engagement-deferred
- price-not-negotiated
- indication-decision-unsupported
6.6.C Mitigation
By Month 9 (2027-07-31), commission a health-economic model from an independent HTA specialist (or an HTA body itself) covering: (a) eligible ALS population (incident and prevalent, by country and disease stage) from a validated epidemiological cohort study; (b) daily-life functional outcome versus eye-tracking (define the outcome precisely, e.g. 'home communication independence 4+ hours/day'); (c) QALY impact over device lifetime (15–20 years, estimate), using published ALS-specific quality-of-life weights; (d) cost-effectiveness at EUR 150,000, EUR 120,000 and EUR 100,000 per implant, and at three QALY thresholds (EUR 80,000, EUR 100,000, EUR 120,000 per QALY); (e) payer budget impact if adoption reaches 50%, 80% and 100% of the annual eligible pool. Simultaneously, conduct pre-reimbursement dialogues with HTA bodies in at least 3 member states (e.g. INAMI/RIZIV Belgium, ZiNL Netherlands, HAS France) by Month 15 (2028-01-31), presenting the model and seeking written feedback on evidence gaps, price tolerance and reimbursement likelihood. Establish a Payer Advisory Board by Month 12 (2027-10-31) with representatives from 3–5 member states to co-design the reimbursement-relevant trial endpoints and the health-economic evidence plan. Schedule a hard Indication Pivot Decision by Month 24 (2028-10-31) (not Month 78 = 2033-04-30): if eligible ALS pool + payer feedback show <300 annual implants and <EUR 100,000 per-QALY threshold, pre-register the move to spinal cord injury and requote the commercial case, timeline and budget with the updated indication. This decision must be published to investors and financiers as a condition of release of T1 funding.
6.6.D Consequence
At Month 24 (2028-10-31) feasibility verdict, the plan cannot credibly answer whether 1,050–5,000 implants is realistic or whether payers will reimburse at EUR 150,000. If EU HTA bodies set the cost-effectiveness bar at EUR 100,000 per QALY (below the proposed ceiling), the device becomes uneconomical unless daily-life benefit is extraordinary, forcing a price renegotiation or a switch to spinal cord injury (larger pool but weaker benefit case). If eligible ALS population is <400 per year (estimate), cumulative implants will be <200 by Month 180 (2041-10-31), guaranteeing insolvency and triggering the Implanted-Patient Continuity Covenant (Decision 8) default, requiring the trust to absorb all long-term support. Investors will demand a reforecast at Months 24 and 36 (2028-10-31, 2029-10-31) gates, and if population and reimbursement remain unvalidated, T2/T3 tranches will be withheld.
6.6.E Root Cause
Decision 3 frames the indication choice as deferrable to Month 78 (2033-04-30) kill gate (Option 3: 'Start with a small ALS communication first-in-human cohort to establish safety and bandwidth, then pre-register a pivotal indication decision at a kill gate'). This appears prudent (keep optionality), but it allows the current plan to be built on a therapeutic population (ALS) whose size and reimbursement prospects are not yet confirmed. No HTA or payer body has been approached; no health-economic model linking daily-life benefit to QALY has been drafted. The plan budgets EUR 500k–€1M for payer engagement in the early analysis (pre-project assessment) but does not schedule it or assign accountability.
7 Expert: Implanting-Centre Network and Surgical Throughput Planner (Multi-Site Coordination, Surgeon Training, Capacity Forecasting)
Knowledge: Neurosurgical centre network design and proctoring, surgeon credentialing and training curricula, operating-theatre capacity and surgical case-load planning, multi-centre clinical-trial coordination and site activation, travel burden and patient equity
Why: The plan targets about 1,050 cumulative implants by Month 180 (2041-10-31) across about 25 centres at a steady-state rate of about 300 per year (about 12 per centre, estimate). The expert validates the Implanting-Centre Network and Surgical Throughput Ceiling decision: whether concentrating on expert centres (high quality, low volume, high patient travel) or distributing to regional hospitals (lower quality risk, higher volume, lower travel) is the right trade-off, and whether the 25-centre footprint and 12 implants per centre per year are realistic given surgeon learning curves, credentialing burden and the small eligible population (500-1,400 per year, estimate). They also assess how the centre network feeds the pivotal indication gate (Month 78 = 2033-04-30): if ALS-only is chosen, can 25 centres stay economically viable, or does the network shrink?
What: Develop a centre-selection strategy and recruitment timeline by Month 12 (2027-10-31): which EU member states, which hospitals, which neurosurgical departments, and what commitment milestones? Design a surgeon credentialing curriculum (duration, proctoring, case minimums, knowledge assessments) by Month 18 (2028-04-30), aligned with implanting-centre network expansion. Forecast annual implant capacity per centre, accounting for learning curves and theatre availability, and validate that the 25-centre, 300-implants-per-year steady-state at Month 168-180 (2040-10-31 to 2041-10-31) is achievable given the eligible ALS pool of 500-1,400 per year. Model the financial and operational implications of narrowing to ALS-only (if the Month 78 = 2033-04-30 gate chooses it) on centre viability and recruitment sustainability.
Skills: Neurosurgical operating-theatre scheduling and throughput optimisation, surgeon training programme design and competency assessment, multi-centre clinical-trial site activation and coordination, learning-curve analysis and case-volume minimums, health-equity and travel-burden assessment, centre economics and case-based reimbursement forecasting
Search: neurosurgical centre network multi-site coordination, surgeon training credentialing implantable devices, operating theatre capacity planning implants, clinical trial site activation and monitoring
7.1 Primary Actions
- Secure written letters of intent from imec, KU Leuven and UZ Leuven by Month 3.5 (2027-02-15), specifying resource commitments (cleanroom hours, staff FTE, patient cohort access) and partnership scope; simultaneously activate fallback sites (UMC Utrecht, Ghent/Antwerp, CEA-Leti/Clinatec) with backup LOIs by Month 3 (2027-01-31) and formal engagement calls by Month 0.5 (2026-11-15). If any Leuven partner declines, trigger fallback activation within 48 hours and compress fallback site qualification to Month 6 (2027-04-30).
- Draft a binding Term Sheet template with explicit gate conditions for T0, T1, T2 release and investor signatures by Month 1.5 (2026-12-15); define automatic deferral or stop triggers at each gate (G1 Month 12 = 2027-10-31, G3 Month 24 = 2028-10-31, G4 Month 36 = 2029-10-31) and write a published wind-down plan (IP preservation, patient-reserve payout, ethics notification) filed with the notified body by Month 3 (2027-01-31). At each gate, publish a binary go/no-go verdict within 2 business days; a missed gate triggers automatic 12-month deferral unless a written investor commitment to extend is signed within 5 business days.
- Commission an independent preclinical evidence review (thin-film surface and penetrating cortical signal stability from published and proprietary datasets) by Month 3 (2027-01-31), with results by Month 9 (2027-07-31); use the 10th and 50th percentiles of signal-retention distributions to set evidence-based kill thresholds (replacing the proposed 20% animal and 25% human thresholds); if thin-film systematically fails the 25% threshold in the literature, trigger an architecture-choice gate at Month 12 (2027-10-31) rather than Month 18 (2028-04-30) and pre-register a contingency-switch decision rule by Month 12 (2027-10-31) with a switch-go decision by Month 48 (2030-10-31). Issue a preliminary feasibility verdict on signal stability by Month 15 (2028-01-31).
- Recruit the 10-core MVP team (founding neurosurgeon, two senior neuroengineers, regulatory, QA, supply-chain, clinical-trials, finance leads) by Month 6 (2027-04-30), with signed contracts including 3-year and 5-year retention bonuses (EUR 20k each) to offset the 20–30% EU pay gap versus US competitors; cross-train a deputy for each role by Month 10 (2027-08-31) and establish weekly all-hands and quarterly retreats by Month 1 (2026-11-30) to sustain cohesion.
- Close EUR 5M in seed funding by Month 3 (2027-01-31) and T0 (floor EUR 10M, target EUR 15M) by Month 12 (2027-10-31), with written source commitments and drawdown schedules published to investors and the Commission by Month 2 (2026-12-31); define three hard financing gates with automatic deferral if missed, backed by technical milestones (Leuven partnerships, notified-body pre-submission, 50% FTO complete by Month 12 = 2027-10-31). Allocate T0 to team (EUR 2.5–3.5M), chronic animal studies (EUR 1.5–2M), FTO and regulatory pre-submission (EUR 0.5–1M), and contingency (EUR 0.5–1M).
7.2 Secondary Actions
- Contact candidate notified bodies (TÜV Süd, BSI, DEKRA Assure) by Month 1 (2026-11-30) and schedule pre-submission meetings for Month 6 (2027-04-30), presenting architecture overview, preclinical programme, clinical-investigation plan, decoder version lock and ISO 13485 roadmap; request written feedback on evidence gaps and regulatory timeline by Month 6.5 (2027-05-15). Request Commission scientific advice by Month 12 (2027-10-31).
- Recruit an independent consent advocate (clinical ethicist or patient representative) by Month 1.5 (2026-12-15) and establish an independent Data Monitoring Committee (5+ members, patient representative, binding pause authority) by Month 6 (2027-04-30); pre-register quantitative safety kill gates (SAE thresholds, signal-stability floors, bandwidth ceilings) by Month 4 (2027-02-28) and obtain ethics-committee pre-approval of consent protocol by Month 6 (2027-04-30).
- Initiate the design-history file (DHF) by Month 1 (2026-11-30) with a designated DHF custodian and secure, versioned repository; establish ISO 13485 QMS charter by Month 3.5 (2027-02-15) and hire or contract a QA Manager with MDR experience by Month 2 (2026-12-31); implement formal change-control process by Month 9 (2027-07-31) and request preliminary QMS audit from the notified body by Month 18 (2028-04-30).
- Commission a freedom-to-operate (FTO) analysis covering all three architecture candidates (thin-film surface, penetrating cortical, endovascular) by Month 3 (2027-01-31), with results by Month 9 (2027-07-31); use FTO scores alongside surgical-risk, chronic-stability and bandwidth modelling to ratify the architecture by Month 18 (2028-04-30); run sovereignty diligence on every licence target using the three EU-control tests (ownership, maintainability without foreign consent, export-control freedom) before pricing options. Publish the upstream-dependency register by Month 12 (2027-10-31).
- Sign MOUs with two competing EU chip suppliers and two substrate suppliers by Month 12 (2027-10-31), with technical datasheets, supply-chain audits and export-control certificates; qualify the second chip and substrate supplier (manufacturing trials, quality testing) by Month 30 (2029-04-30). If multi-state foundry commitments are not confirmed by Month 18 (2028-04-30), activate the dual-sourcing route and fund pilot-line-to-validated-production transfer (EUR 10M in T1, EUR 25M in T2) before design lock at Month 72 (2032-10-31).
- Commission epidemiological analysis of the eligible ALS-communication pool (age, disease stage, life expectancy, consent capacity) by Month 6 (2027-04-30), with results by Month 12 (2027-10-31); if the eligible pool is <300 per year, pre-register a Month 24 (2028-10-31) scope gate choosing between ALS-only (with smaller scale), spinal-cord-injury pivot (with harder benefit comparison), or programme deferral. Begin payer pre-engagement dialogues with 3+ member-state HTA bodies (France, Belgium, Netherlands) by Month 9 (2027-07-31), testing willingness to pay at EUR 150,000 per implant and cost-effectiveness ceiling of EUR 120,000 per QALY.
- Establish a patient-protection trust by Month 6 (2027-04-30) with EUR 5M floor before FIH, funded by EUR 30,000 per implant at implantation, held outside investor reach; draft a wind-down and explantation-support plan by Month 3 (2027-01-31) and file it with the notified body and ethics committees by Month 6 (2027-04-30).
- Commission a Leuven climate-resilience assessment (heatwaves above 35°C, drought, flooding, power outages, water supply, thermal stability of wearable hardware) by Month 1 (2026-11-30), with results by Month 9 (2027-07-31); by Month 18 (2028-04-30), install backup power, redundant cooling and a second-site compute replica; design the wearable for ≥12 hours battery life with passive cooling and automatic throttling, tested in climate chambers at 35–40°C by Month 24 (2028-10-31).
- Establish a separate legal entity for human-AI enhancement research by Month 6 (2027-04-30) with a documented firewall governing IP and hardware sharing; publish a one-page intended-use statement for the medical device clarifying that enhancement is not a current application; commission an independent external ethics and public-engagement board to review enhancement communications by Month 9 (2027-07-31).
7.3 Follow Up Consultation
At Month 3 (2027-01-31), confirm Leuven partnerships in writing or activate fallback sites; review T0 funding closes and team recruitment progress. At Month 6 (2027-04-30), verify notified-body pre-submission meeting occurred, DMC is seated with published kill gates, and patient-protection trust is funded. At Month 12 (2027-10-31), confirm G1 financing committed or trigger automatic deferral; review FTO results, EU-control sovereignty tests on licence targets, and supplier MOU signatures. At Month 15 (2028-01-31), publish preliminary feasibility verdict on signal stability (evidence-based kill thresholds). At Month 24 (2028-10-31), issue final six-part feasibility verdict on scientific, engineering, clinical, regulatory, supply-chain and financing feasibility; confirm G3 funding or stop the programme. If any gate is missed, convene investors, the Commission and ethics advisors within 5 business days to execute the pre-registered fallback (deferral, switch or wind-down).
7.4.A Issue - Leuven partnerships remain uncommitted, risking cascading schedule slip and cost overrun
The plan assumes imec, KU Leuven and UZ Leuven will partner but none has signed a commitment. The fallback pathway (Ghent, Utrecht, Grenoble) is documented but adds 3–6 months and up to EUR 2M. This is the single point of failure for cleanroom access, surgical training, ALS patient recruitment and preclinical facilities. No signed letters of intent exist by Month 0 (2026-10-31). Without written commitments by Month 3 (2027-01-31), the 10-month founding-team build and architecture lock at Month 18 (2028-04-30) both slip, compressing the FIH window and risking a competitive-response gate activation.
7.4.B Tags
- partnership_risk
- schedule_compression
- single_point_of_failure
- contingency_not_activated
7.4.C Mitigation
Immediate action (within 2 weeks): the founding team must contact imec, KU Leuven and UZ Leuven with a one-page partnership scope (cleanroom hours, staff FTE, patient cohort access, decision deadline). Request LOIs by Month 3.5 (2027-02-15). In parallel, activate fallback talks with UMC Utrecht, Ghent/Antwerp and CEA-Leti by Month 0.5 (2026-11-15), securing backup LOIs by Month 3 (2027-01-31). If any Leuven institution declines, trigger fallback activation within 48 hours and compress fallback site qualification to 8 weeks (Month 6 = 2027-04-30), updating FIH and architecture-lock milestones. Publish a written partnership status report to investors and the Commission by Month 2 (2026-12-31).
7.4.D Consequence
If partnerships are not confirmed by Month 3 (2027-01-31), the MVPfunding (EUR 15M) cannot be committed reliably, because the cost and timing of a fallback site are uncertain. FIH slips to Month 66–72 (2032-04-30 to 2032-10-31), pushing CE marking beyond Month 140 (2038-06-30). The 5,000-implant target becomes even less credible. Investor and Commission confidence erodes, making tranches G1 (Month 12 = 2027-10-31) and G3 (Month 24 = 2028-10-31) impossible to close.
7.4.E Root Cause
The plan was written before founder recruitment and partnership outreach began. No scoping calls or LOI templates exist. The Leuven partners were listed as 'none of which has committed', signalling this was a known gap. The fallback pathway was documented but never formally triggered or tested; no backup site has received a formal partnership inquiry.
7.5.A Issue - Financing gates are declared but not gated: no written fallback or automatic deferral mechanism binds investors or stops the programme
The plan defines three hard gates (G1 Month 12 = 2027-10-31, G3 Month 24 = 2028-10-31, G4 Month 36 = 2029-10-31, 2028-10-31, 2029-10-31) with written verdicts within 5 business days. However, no investor commitment letter templates, no automatic deferral triggers and no stop mechanism exist. The EUR 15M T0 'floor' is termed 'proposed' yet drives the entire MVP. If a gate is missed, the plan says 'stop, defer or switch' but does not specify who decides, what the alternative is, or what happens to implanted patients. This vagueness converts gates from credible kill switches into rhetorical aspirations.
7.5.B Tags
- financing_gate_not_binding
- no_stop_mechanism
- investor_commitment_undefined
- ambiguous_fallback
7.5.C Mitigation
By Month 1.5 (2026-12-15), draft a Term Sheet template with explicit gate conditions: 'T0 release is contingent on (a) Leuven partnerships LOI by Month 3 (2027-01-31), (b) DMC established by Month 6 (2027-04-30), (c) notified-body pre-submission by Month 6 (2027-04-30), (d) FTO review 50% complete by Month 12 (2027-10-31).' At each gate date, publish a go/no-go verdict to investors, the Commission and a public stakeholder list within 2 business days. If G1 (Month 12 = 2027-10-31) is missed, automatically halt non-critical hiring and defer Months 13–24 (2027-11-30 to 2028-10-31) spend within 48 hours unless a written extension commitment is signed by the lead investor. Write a wind-down plan (preserve IP, notify ethics committees, patient-reserve payout schedule) and file it with the notified body by Month 3 (2027-01-31) so it is immediately executable.
7.5.D Consequence
If gates lack binding teeth, the board will negotiate past each deadline rather than stop. Implanted patients will accumulate while the financing situation deteriorates, creating moral hazard and destroying the 'hard requirement' credibility that regulators and ethics committees require. The plan's claim to be 'gated' becomes an empty gesture. Investor trust evaporates when the first gate is missed and the company continues anyway.
7.5.E Root Cause
Empty
7.6.A Issue - Feasibility verdict for the original mission is deferred to Month 24 (2028-10-31), but kill-gate thresholds for signal stability and bandwidth are unvalidated and may be unrealistic
The plan pre-registers kill thresholds: 20% signal loss blocks FIH, 25% blocks continuing in humans (proposed), 5 bits/s in 50% of users triggers review (proposed). No published preclinical data is cited to justify these targets. Thin-film surface architectures in the literature show 30–50% signal decay at 12 months; penetrating arrays show better stability but higher surgical risk. The EUR 30M architecture-switch contingency is budgeted but switching at Month 72 (2032-10-31) compresses the pivotal-lock window to 12 months (Monster 72–84 = 2032-10-31 to 2033-10-31), risking CE-marking slip to Month 156 (2039-10-31). If signal stability is the real blocker, the plan needs to justify the 20% and 25% thresholds with evidence before FIH, not guess them post-hoc.
7.6.B Tags
- kill_gate_validation_missing
- signal_stability_risk
- architecture_switch_contingency_tight
- feasibility_verdict_deferred
7.6.C Mitigation
By Month 9 (2027-07-31), commission an independent preclinical review of thin-film surface and penetrating cortical chronicity from published data and unpublished proprietary datasets (imec, Clinatec, UMC Utrecht). Calculate signal-retention percentiles and 12-month decay distributions; set kill thresholds at the 10th percentile (pessimistic) and 50th percentile (central) to anchor them in evidence. If thin-film signal loss systematically exceeds 25% at 12 months in the literature, revise the threshold to 35% or trigger an earlier architecture-choice gate at Month 12 (2027-10-31) rather than Month 18 (2028-04-30). Document the 20% animal and 25% human thresholds as informed by this review, not guesses. Pre-register the contingency-switch decision rule: e.g. 'if mean signal retention at Month 36 (2029-10-31) of the first 3 animals is <50%, trigger Month 42 (2030-04-30) architecture review, with switch decision by Month 48 (2030-10-31).' Issue a preliminary feasibility verdict on signal stability by Month 15 (2028-01-31).
7.6.D Consequence
If the 20% and 25% thresholds are not evidence-based and are crossed in early animals, the plan will face a forced architecture switch at Month 48–60 (2030-10-31 to 2031-10-31), precisely when FIH should be imminent. This compresses the pivotal-trial preparation window to under 24 months, making CE marking by Month 132 (2037-10-31) impossible and sliding it to Month 150+ (2039-04-30+). The switching cost (EUR 30M, 12–24 months) will blow through the free contingency. The feasibility verdict at Month 24 (2028-10-31) will be forced into 'not credibly achievable' rather than 'achievable only with changes', and investors will have a credible reason to walk away.
7.6.E Root Cause
Empty
8 Expert: Patient Protection, Continuity and Reserves Governance (Explantation, Device Lifetime, Company-Failure Scenarios)
Knowledge: Medical-device explantation and failure-management protocols, actuarial reserve sizing and trust administration, patient-protection covenants and step-in agreements, long-term device follow-up registries and cybersecurity, company insolvency scenarios and patient-support continuity
Why: The plan designates patient protection as a hard requirement, with a ring-fenced trust of EUR 30,000 per implant (estimate), totalling about EUR 31.5M for the central 1,050 implants by Month 180 (2041-10-31). The expert validates the Implanted-Patient Continuity Covenant decision: whether the reserve is adequate to cover follow-up, servicing, explantation and data access for the expected device lifetime (to be estimated), whether a step-in agreement with a successor manufacturer or hospital network is credible, and whether the device design (open external interfaces, simple explantation path) makes patient support durable if the company fails. They also assess whether the trust structure (funded at implantation, held outside investor reach) is legally and operationally sound.
What: Commission an actuarial analysis by Month 9 (2027-07-31) to size the per-implant reserve: estimate device lifetime (central case 10-15 years, estimate), annual follow-up costs (clinic visits, imaging, wireless charging, software updates), explantation costs, and revision/replacement rates. Establish the trust structure and governance by Month 6 (2027-04-30), with independent trustees, a published reserve policy, and auditing. Design explantation and device-failure protocols by Month 24 (2028-10-31), with surgical simplicity validated in models or cadavers. Negotiate or draft a step-in agreement with a second manufacturer or EU hospital consortium by Month 36 (2029-10-31), and test the handover with a named successor. Define cybersecurity and data-access continuity: what happens to neural data if the company fails, and how long are records kept?
Skills: Medical-device failure analysis and recall management, explantation-procedure design and surgical validation, actuarial reserve sizing and long-term forecasting, trust and escrow administration under European law, step-in agreements and succession planning, patient-registry governance and lifelong follow-up protocols, cybersecurity and data-destruction policies
Search: medical device explantation protocol failure management, patient protection reserve actuarial sizing, device lifetime long-term follow-up registry, company insolvency patient support continuity
8.1 Primary Actions
- By Month 2 (2026-12-31): Retain an independent trustee for the patient-protection reserve with medical-device escrow experience; commission a device-lifetime actuarial study; draft a trust deed specifying trustee duties, successor-access rights, drawdown triggers and investment policy; share the framework with the notified body and ethics committees for pre-submission feedback.
- By Month 9 (2027-07-31): Define explantation as a non-negotiable design input; specify reversibility criteria (electrode recovery, tissue safety, surgical time and complication targets); mandate a dedicated explantation-validation cohort in chronic-animal studies (3-5 animals, 14-24 months); publish an independent surgeon-validated explantation protocol; fund explantation from the patient-protection reserve (cost estimate EUR 5,000–EUR 15,000 per procedure); include explantation outcomes (tissue damage, surgical time, complication rates) as a kill-gate measure.
- By Month 9 (2027-07-31): Publish a Decoder Lifecycle Policy defining patient data ownership, AI model IP residency (EU-held), update classification (minor autonomous, major notified-body reviewed), Decoder Registry transparency, post-CE-marking audit logging and patient data-export rights; conduct a cybersecurity assessment covering cloud access, encryption, data isolation and incident response; if decoder updates exceed 12 per year, commit to re-audit every 6 months.
- By Month 3 (2027-01-31): Secure written letters of intent from imec, KU Leuven and UZ Leuven specifying resource commitments, patient-access guarantees and decision timelines; simultaneously activate parallel backup-site engagement with UMC Utrecht, Ghent/Antwerp and CEA-Leti/Clinatec; if any primary site declines, activate the fallback immediately and compress site qualification to 12 weeks (decision by Month 6 = 2027-04-30).
- By Month 6 (2027-04-30): Establish an independent Data Monitoring Committee (5+ members: two external statisticians/epidemiologists, one patient representative, one ethics specialist, one external clinician) with binding authority to pause, modify or stop implantation; require monthly safety review (company submits SAEs, signal-loss and bandwidth data by the 5th; DMC meets within 10 days and publishes findings); contractually bind investor tranches (Series B at Month 24 = 2028-10-31) to DMC safety clearance; pre-register quantitative kill-gate thresholds before first-in-human.
- By Month 12 (2027-10-31): Publish a written Upstream Supplier Dependency Register listing all non-EU tools, EDA software, fab equipment and components, with explicit mitigation plans; sign MOUs with two competing EU chip suppliers and two substrate suppliers, specifying exclusivity, second-source qualification timelines (Month 30 = 2029-04-30) and post-CE-marking service obligations; pass all components against the three EU-control tests (EU ownership, maintainability without foreign consent, export-control freedom) by Month 72 (2032-10-31).
8.2 Secondary Actions
- By Month 3 (2027-01-31): Commission a Leuven site resilience assessment covering multi-day heatwaves above 35°C, drought, flooding, power outages, water cooling and thermal stability of wearables and patients with impaired thermoregulation; identify dependencies and mitigation costs; report findings to the board and environmental-risk committee.
- By Month 6 (2027-04-30): Design and validate a yes/no communication-validation protocol for locked-in participants, using their best available method (eye-tracking, blink, assistive device) with at least two independent validators and video recording; submit to three ethics committees (Leuven, Utrecht, Grenoble) for pre-review by Month 3 (2027-01-31); incorporate feedback and obtain written ethics approval by Month 5 (2027-03-31).
- By Month 9 (2027-07-31): Complete a bottom-up eligible-population epidemiology study covering ALS incidence, progression rates, survival, attrition and centre/theatre capacity across 3+ member states; compare the 1,050 cumulative implants (central estimate) against eligible pool and annual surgical throughput; identify gaps and recommend indication-pivot triggers for the Month 24 (2028-10-31) scope gate.
- By Month 12 (2027-10-31): Initiate payer and HTA pre-reimbursement dialogues in at least three member-state healthcare systems; present health-economic model (EUR 150,000 device cost, EUR 120,000 per QALY ceiling proposal) and ask: (a) what daily-life functional outcome vs. eye-tracking would win reimbursement?; (b) what evidence package is required?; (c) what price would payers accept?; (d) would outcome-based contracts be acceptable?; document findings and incorporate into pivotal-trial endpoint design.
- By Month 15 (2028-01-31): Conduct freedom-to-operate analysis against known Neuralink, Synchron, university and patent-office BCI patents for the primary architecture (thin-film surface) and the penetrating fallback; identify blocking claims and design-around costs; if blocking claims exist, budget licence or challenge costs in T1 or T2 tranches; document this analysis as a sovereignty test input.
- By Month 18 (2028-04-30): Request a preliminary ISO 13485 QMS audit from the selected notified body; incorporate findings into the final QMS by Month 24 (2028-10-31); ensure design-history file, change control, training records and supplier audit procedures are audit-ready before design lock.
- By Month 24 (2028-10-31): Publish a comprehensive six-part feasibility verdict: (1) scientific (chronic stability, bandwidth, safety); (2) engineering (architecture performance, manufacturing transfer); (3) clinical (indication choice, eligible population, recruitment pace); (4) regulatory (notified-body alignment, AI Act scope); (5) supply-chain (foundry pathway, sovereignty tests passed or remaining); (6) financing (tranches committed, revenue vs. opex gap, break-even analysis). Distinguish achieved targets from missed targets and revised assumptions; present this verdict to investors, member states and the Commission within 5 business days of Month 24 (2028-10-31).
- By Month 30 (2029-04-30): Qualify and validate a second EU chip supplier and a second substrate supplier, with full process transfer, yield data and quality audits; ensure both meet the three EU-control tests and can sustain production if the first supplier fails or exits.
- By Month 36 (2029-10-31): If a US or Chinese device achieves CE marking in the EU during Months 36–60 (2029-10-31 to 2031-10-31), activate a competitive-response gate: assess whether the EU device still has a clinical or trust advantage (EU data residency, local service, lifelong support, explantation commitment); decide whether to accelerate FIH, adjust the payer pitch or pivot indication to a segment less vulnerable to US competition.
8.3 Follow Up Consultation
At the Month 3 (2027-01-31) check-in, review: (1) trustee retention, actuarial study and trust-deed draft status; (2) explantation-design criteria and chronic-animal study integration; (3) Decoder Lifecycle Policy publication and cybersecurity assessment scope; (4) Leuven partner LOI status and fallback-site activation triggers; (5) Data Monitoring Committee recruitment and kill-gate pre-registration completion; (6) Upstream Supplier Register and first-supplier MOU status. By Month 9 (2027-07-31), expect: explantation protocol validation, eligible-population epidemiology, payer-dialogue readiness and freedom-to-operate analysis outline. By Month 12 (2027-10-31), closure of the first USD 15M financing gate (T0) and notified-body pre-submission meeting. The Month 24 (2028-10-31) verdict is the most critical output; withhold Series B capital release until all six feasibility domains are presented. Thereafter, quarterly reviews should track compliance with kill gates, consent governance, decoder change-control and the Decoder Registry; annual audits should verify trustee performance, reserve adequacy and patient-data stewardship.
8.4.A Issue - Patient-Protection Reserve Governance Is Incomplete and Critically Under-Specified
The plan mandates a EUR 30,000-per-implant reserve by Month 6 (2027-04-30) with a EUR 5M floor before first-in-human, but does not specify: (1) who holds and administers the trust (independent trustee, hospital, notified body, regulator?); (2) what triggers drawdown (device failure, company insolvency, explantation, long-term support); (3) how device lifetime and actuarial adequacy are assessed; (4) whether reserves are invested, and what returns are assumed; (5) how successor parties gain access if the company fails; (6) what happens to surplus reserves after patient death or explantation. Without these details, the reserve is legally ambiguous and may not survive regulatory or payer scrutiny. A failed or acquired company could claim the reserves; patients could be abandoned if the trustee is not bound to explicit support obligations.
8.4.B Tags
- patient_protection_governance_gap
- trust_administration_undefined
- successor_handover_missing
- actuarial_basis_absent
- legal_enforceability_risk
8.4.C Mitigation
By Month 2 (2026-12-31), retain an independent trustee (e.g., a law firm, bank trust department or EU patient-protection foundation) with experience in medical-device escrow. Draft a trust deed specifying: (a) trustee duties to support implanted patients for device lifetime (to be actuarially modeled by Month 9 = 2027-07-31); (b) drawdown triggers (device failure, company insolvency, patient death, explantation); (c) successor-access rights (named fallback manufacturers, hospitals, the notified body); (d) investment policy and return assumptions; (e) annual audit and reporting to a patient advocate and regulator. Have the notified body review the deed as part of pre-submission (Month 6 = 2027-04-30 meeting). Publish the trust framework by Month 6 (2027-04-30) to investors and ethics committees.
8.4.D Consequence
Without explicit trust governance, the reserve may be legally challenged as company assets, freezing patient support if the company is acquired or enters insolvency. Regulators and ethics committees may block first-in-human without clear successor-access rights and actuarial backing. Payers may refuse reimbursement if patient protection is ambiguous. Post-CE-marking, patients implanted under an unclear covenant could sue if support is withdrawn or insufficient, damaging the company and the programme.
8.4.E Root Cause
The plan treats the patient-protection reserve as a financial gate-keeping item rather than a binding legal obligation with a defined trustee and successor-access mechanism. The actuarial basis (device lifetime, failure modes, attrition, explantation rates) is missing, so the EUR 30,000 figure is unsupported. No precedent is cited for how other implantable-device companies or EU regulators structure such trusts.
8.5.A Issue - Explantation Pathway, Device Design and Surgical Recovery Are Not Integrated Into Architecture or Kill Gates
The plan mentions explantation as part of patient continuity but does not specify: (1) design criteria for reversibility (electrode material, encapsulation, attachment strength, recovery protocols); (2) surgical risk, tissue scarring and recovery timeline; (3) who implants and who explants (same team, certified specialist, any trained neurosurgeon?); (4) what happens to explanted devices (data extraction, device disposal, tissue analysis); (5) how explantation readiness is validated preclinically; (6) when a patient can request explantation (end of trial, device failure, patient choice, adverse event); (7) cost attribution (company, payer, patient, trust). Omitting explantation from the architecture choice and kill gates means the device could be designed for implantation only, with no reversible exit. This violates informed consent and patient autonomy.
8.5.B Tags
- explantation_design_missing
- reversibility_not_specified
- surgical_risk_undefined
- informed_consent_gap
- device_lifecycle_incomplete
8.5.C Mitigation
By Month 9 (2027-07-31), define explantation requirements as design inputs: (a) electrodes must be retrievable with <10% tissue damage; encapsulation must not fuse to dura or cortex; (b) surgical explantation time must be <4 hours, with complication rates <5%; (c) tissue and electrode recovery must preserve histology and electrolyte balance for post-mortem or clinical analysis. Test reversibility in chronic-animal studies (14-24 months); include a dedicated explantation cohort (3-5 animals) with full surgical and tissue assessment. By Month 18 (2028-04-30), publish an explantation protocol with independent surgeon validation. Include explantation as a mandatory kill-gate measure: if >2 explantations in the first 20 implants show significant tissue damage or >2-hour surgical time, trigger design review. Fund explantation from the patient-protection reserve; cost estimate EUR 5,000-EUR 15,000 per procedure.
8.5.D Consequence
Implanted patients facing device failure, adverse events or their own choice to exit may be trapped with a non-reversible device, unable to pursue alternative treatments or manage complications. Surgeons may lack training or capability to explant safely, leading to tissue damage or death. Ethics committees and payers may refuse to approve a trial if explantation is not credibly possible. If explantation complications are discovered late (Year 5+), the company faces product-liability claims and regulatory action. The plan could be forced to abort implantation or offer emergency explantations without proper protocol.
8.5.E Root Cause
The plan separates device design, surgical procedure, patient consent and the continuity covenant, treating explantation as a downstream support obligation rather than a design requirement. Neurotech precedent (Clinatec, UMC Utrecht) shows that surface and endovascular devices are more reversible than penetrating designs, but this trade-off is not explicit in the architecture choice. The plan defers indication and architecture decisions to Month 18 (2028-04-30), too late to shape explantation reversibility into the design baseline.
8.6.A Issue - Decoder Lifecycle, Data Ownership and Cloud Dependency Create Unmanaged Sovereignty and Cybersecurity Risks
The plan specifies a 'locked decoder version' for the pivotal trial and version-gated change control for the AI Act, but does not address: (1) who owns the training data (participant neural signals) and models (weights, architecture, features) after CE marking; (2) whether decoder updates after pivotal lock require notified-body approval or can be autonomous; (3) whether patient data or model weights can be shared across patients, centres or member states; (4) whether the decoder runs on-device, on a body-worn processor, or on EU-hosted cloud; (5) what happens if the company is acquired and the acquirer is non-EU or is subject to export controls; (6) whether patients retain the right to their neural data or to export it; (7) how long neural data is kept and who accesses it for AI training. If decoder updates are frequent and the cloud dependency is opaque, patients lose control of their device behaviour and their data becomes a silent asset extraction risk.
8.6.B Tags
- decoder_ownership_undefined
- data_sovereignty_gap
- cloud_dependency_unmanaged
- ai_model_lineage_unclear
- patient_data_rights_absent
- acquisition_scenario_unaddressed
8.6.C Mitigation
By Month 9 (2027-07-31), publish a Decoder Lifecycle Policy specifying: (a) participant neural signals are owned by the patient; the company holds a limited licence for clinical care and pre-registered research only; (b) AI model weights and feature sets are EU-held IP, not sold or transferred outside the EU; (c) decoder updates classified as minor (retraining within locked feature set) are autonomous; major updates (new features, new data sources, architecture swap) require notified-body scientific review; (d) post-CE-marking, all decoder versions and validation results are logged in a publicly queryable Decoder Registry to support transparent audits; (e) if the company is acquired, the acquirer must assume EU data residency and model-lineage obligations as a condition of closing; (f) patients can request data export (neural signals, model weights, performance metrics) within 30 days. Run a cybersecurity assessment by Month 12 (2027-10-31) covering: cloud access controls, encryption, data isolation per patient, audit logging and incident response. If decoder updates occur >12 times per year, conduct independent security re-audit every 6 months.
8.6.D Consequence
Without clear ownership and access rules, patients' neural data could become a training corpus for a company's or acquirer's broader AI models, violating consent and GDPR. Frequent decoder updates without proper validation could degrade performance or introduce unintended biases. An acquirer could relocate AI training to non-EU jurisdiction, failing the sovereignty tests. Regulators could require model revalidation after each update, slowing beneficial improvements. If a cybersecurity breach exposes neural signals, patients face identity theft (neural fingerprinting) or blackmail (mental-state inference). The company could face GDPR fines (up to EUR 20M or 4% of revenue) and patient lawsuits.
8.6.E Root Cause
The plan focuses on decoder change control for the pivotal trial but does not govern the long-term ecosystem after CE marking, when patient numbers grow and continuous AI improvement becomes operationally normal. The sovereignty tests are applied to hardware and manufacturing, not to AI training data and model ownership. No analogous medical-device precedent is cited; the plan assumes a static decoder rather than a learning system.