limen Step 5 — no target selected

PRODUCER CANDIDATE — PENDING INDEPENDENT QA AND INTEGRATOR ACCEPTANCE

CONCEPT STUDY · RESEARCH ONLY · NO HUMAN TESTING · NOT AVAILABLE · NO DEVELOPMENT PROGRAM EXISTS

Executive Summary

• Decision: NO-SELECTION. The accepted Step 5 evidence does not justify choosing a target, route, architecture, device, product, or development program.
• Closest tuple, not a choice. S5R2-T01 asks whether adults with acquired unilateral transradial limb loss could perform vision-occluded contact and release object manipulation using direct natural peripheral-nerve electrical activity (ENG) forward and afferent peripheral-nerve stimulation return.
• Current result: hold. Candidate UNKNOWN; optimized surface muscle electrical activity (EMG) plus external-feedback comparator UNKNOWN; exact complete-match search OPEN. The deterministic result is HOLD_BOTH_UNKNOWN.
• Authority stays narrow. Product supplies a neutral functional frame, A/B/C remain unselected, the Legal, Regulatory and Ethics gate remains independent, the overall project remains REVISION, and release is unauthorized.

The few terms needed

Direct natural ENG is naturally occurring electrical activity recorded from a peripheral nerve. It is not muscle activity, an evoked response, or stimulation artifact.

Surface EMG is muscle electrical activity recorded through skin electrodes. Here it is the removable forward comparator.

Complete route means the forward function, return function, population, endpoint, schedule, and adequate-duration rule considered together. Separate component results cannot be added together and called a complete result.

HOLD means required evidence is missing, mismatched, or insufficient. It is not evidence failure and it is not permission to proceed.

The deferred route and its matched comparator

Candidate research route CNS → peripheral nerve → direct natural ENG → computer command computer → stimulation interface → afferent peripheral nerve → CNS Matched lower-burden comparator removable surface EMG → computer command external vibrotactile or electrocutaneous contact/release feedback → person

This diagram names information routes only. It contains no device layout, biological setting, implant choice, stimulation value, or implementation.

Direct-neural uniqueness fails only at one narrow endpoint

PMID 33725672 reports one participant, simultaneous TIME neural and surface-EMG acquisition, and similar force/velocity decoding across three recording days. That makes a unique direct-neural advantage FAIL at this narrow forward-component endpoint.

It supplies no return path, no prospectively adequate chronic window, and no complete-route advantage. The full-text body was inaccessible in the accepted review, so exact calendar span, adverse-event accounting, funding, and conflicts remain UNKNOWN.

The external comparator raises the bar but does not close it

PMID 26584497 reports five people with transradial limb loss using myoelectric control with discrete external vibrotactile contact/release feedback at home for one month. It is a relevant comparator component.

It is not an eligible candidate-versus-comparator head-to-head complete-route comparison, and the accessible abstract does not supply the numerical result, adverse-event accounting, funding, or conflict detail needed for broader claims.

The complete matched comparison is still missing

No inspected source matches the same population, joint forward-plus-return function, vision-occluded contact/release endpoint, optimized comparator, and prospectively adequate duration. The search result stays OPEN, not proof of absence.

So what: component precedents constrain claims, but they cannot be pooled into selection. Candidate UNKNOWN plus comparator UNKNOWN maps to HOLD_BOTH_UNKNOWN.

Five candidates were routed; none was selected

Two tuples are DEFER and three are HOLD. Zero are selected. The chart counts exact accepted dispositions across all five candidate tuples; it is not an evidence score, probability, priority score, or readiness measure.

Companion count table

The table repeats the chart's exact counts across the same denominator of five candidate tuples. It is not an evidence score, probability, priority score, or readiness measure.

Exact candidate record

• S5R2-T01 — DEFER, closest tuple; not selected: vision-occluded contact/release object manipulation after acquired unilateral transradial limb loss. Candidate UNKNOWN; comparator UNKNOWN; complete-match search OPEN.
• S5R2-T02 — DEFER: reliable computer access for severe paralysis. Lower-burden comparator evidence exists; no complete peripheral candidate arm.
• S5R2-T03 — HOLD, not evidence FAIL: bladder-state monitoring and modulation in neurogenic lower urinary tract dysfunction. Eligible human exact-route comparison OPEN; candidate UNKNOWN.
• S5R2-T04 — HOLD, not evidence FAIL: treatment-refractory post-amputation pain. Eligible bidirectional comparison OPEN; direct-ENG forward function UNKNOWN.
• S5R2-T05 — HOLD, not evidence FAIL: peripheral-nerve recovery monitoring and rehabilitation feedback. Eligible exact candidate/comparator match OPEN; candidate UNKNOWN.

Why unknown plus unknown means hold

• Candidate MEETS plus comparator BELOW can retain only a bounded research hypothesis; it does not establish superiority.
• Comparator MEETS_OR_EXCEEDS makes the unique-function claim FAIL regardless of candidate result.
• Candidate DOES_NOT_MEET independently stops the candidate.
• Candidate UNKNOWN plus comparator UNKNOWN gives HOLD_BOTH_UNKNOWN.
• Any population, route, function, endpoint, schedule, or duration mismatch remains UNKNOWN; another bounded negative search remains OPEN.

Product remains neutral

The accepted Product snapshot is a functional and decision-gate frame only. It defines abstract forward, return, calibration, monitoring, provenance, service, and fail-closed responsibilities. A — conventional electronic, B — biohybrid/living intermediary, and C — combined biological/electronic remain unselected.

So what: Product can translate a separately accepted target later; it cannot manufacture one from component evidence or founder preference.

Legal, ethics and regulatory review remains an independent gate

No intended use, population, architecture family, product category, jurisdiction, regulatory path, patentability position, or FTO conclusion is selected. Consent, privacy, command integrity, reversibility, access, equity, dual-use, classification, and long-term accountability remain independent gates.

So what: scientific evidence cannot waive governance, and governance language cannot upgrade scientific evidence.

Next research work, in order

State: REQUIREMENT. These are dependency-ordered research and governance gates, not claims of completed work.

1. Independently review this report candidate. QA must reproduce the exact 12-file package; the Integrator must separately accept identical bytes before the report can become an accepted explanatory view.
2. Keep the complete-match search open. Reopen only primary or official records that can match population, route, function, endpoint, schedule, duration, denominators, adverse events, revisions, explants, interruptions, category identity, and recalibration burden.
3. Re-adjudicate only eligible evidence. Apply the accepted deterministic table without pooling forward, return, muscle, stimulation-only, sequential, or adjacent-route components.
4. Run the independent Legal, Regulatory and Ethics gate on the same fixed record. Affected-community, clinical, disability-rights, bioethics, privacy, cybersecurity, regulatory, and jurisdiction-specific review may narrow or stop the question.
5. Translate only after selection evidence exists. A separately accepted Science target, matched comparator result, Product translation, QA, Integrator acceptance, and explicit user authority would all be required before any later scope.

Blocked engineering: no blueprint, device specification, architecture, component, synthetic vector, simulator, software, bench, build, test, procurement, fabrication, implantation, biology, animal work, or human work is authorized.

What could change the decision?

Only eligible evidence under the frozen rule: the candidate must MEET the joint endpoint and the optimized matched comparator must be BELOW over the same prospectively adequate window. Comparator MEETS_OR_EXCEEDS defeats uniqueness; candidate DOES_NOT_MEET stops the candidate; mismatched or missing evidence keeps the result on hold.

Acceptance state and limits

The Product, Science, and Legal, Regulatory and Ethics input snapshots are accepted only within their exact scopes and hashes. This new founder report is not yet accepted.

PRODUCER CANDIDATE — PENDING INDEPENDENT QA AND INTEGRATOR ACCEPTANCE. The overall project remains REVISION and release remains unauthorized. Nothing in this report establishes clinical benefit, safety, efficacy, feasibility, novelty, patentability, FTO, universality, regulatory classification, readiness, development authority, or availability.

CONCEPT STUDY · RESEARCH ONLY · NO HUMAN TESTING · NOT AVAILABLE · NO DEVELOPMENT PROGRAM EXISTS