idea-002 · feasibility deck
Every slide states a condition that would have to be true, then reports where it stands using the candidate file's own claim and its own confidence tag. A deck never upgrades a tag, invents a number, or recommends anything, and there is no ask slide.
The chair: the engineers who have to make an adhesive hold vacuum on fur, on a limb, on an animal that moves — and a disposable pump that keeps it there for the whole course.
How to read this: every slide is a condition, not a conclusion. A bracketed claim reference points into knowledge-base/candidates/idea-002.md: the number is the claim's position in its ## Claims list, and the tag is the one that claim carries there, copied and never adjusted here. A [no claim] marker means nothing in the file speaks to the condition. See docs/deck-spec.md.
What changed since the 2026-09-24 build: the pump's design envelope went to the Verifier and came back unconfirmed on every line, for want of an instrument rather than on contrary evidence — duty cycle, control band, battery and RF (radiofrequency) regimes, cell-safety and 60601 scope claim 65unconfirmed through claim 69unconfirmed. The battery/RF line split on the regulation text: the lithium-transport half matched, and the FCC half's "no exemption available" was left unestablished with at least two open routes claim 67unconfirmed. The human NPWT (negative-pressure wound therapy) special controls moved to verified, including the scope text that confines them to human-use devices claim 99verified. This rebuild appended claims 152 and 153. Rebuilt from the file, not patched.
Would have to be true: A disposable battery pump would have to hold approximately -80 mmHg (millimetres of mercury) at the dressing, through a superabsorbent canister-free dressing, across a fur-compatible drape and a limb-conforming interface, on open, contaminated or grafted canine and equine distal-limb wounds — absorbing all of the exudate itself — for the whole treatment course.
Where it stands: The requirements now exist as numbers and none has been measured. Seal: at least 72 hours at approximately -80 mmHg under normal ambulation claim 28unconfirmed. Energy: mean pump ON-time at or under 7% of any 24 hours, the figure a 6 Wh primary-cell budget and a 7-day single-use therapy are sized against claim 65unconfirmed. Control: -80 mmHg within +/-10 mmHg measured at the dressing port, with an independent hardware limit stopping anything more negative than -110 mmHg under any single fault claim 66unconfirmed. No claim states a leak-rate limit of its own; the nearest the file comes is that duty figure, set for a dressing held on an ambulatory limb claim 65unconfirmed. Two requirements this rebuild added because nothing in the file carried them: the dressing absorbs an open or contaminated wound's exudate across each wear interval without saturating claim 152unverified, and the -80 mmHg setpoint sits within the pressure range of the favourable veterinary results claim 153unverified. Course length has a tension against the 7-day sizing: mean NPWT application in the 42-horse series was 11.5 days claim 36unconfirmed. The file still records a separate pump design-input document as unconfirmed claim 60unconfirmed.
What would settle it: Instrumented bench measurement against these numbers on a limb model with a calibrated leak; exudate-loading tests on the dressing at the fluid volumes an open distal-limb wound produces; a human reading of the favourable studies' methods for the pressures they used.
If it's false: There is no spec a build can pass, and — unusually in this repo — no regulator will supply one (slide 3).
Would have to be true: NPWT's benefit would have to survive translation into dogs and horses, onto the distal limb, into the canister-free single-use format, at approximately -80 mmHg — on open and contaminated wounds, not only closed incisions.
Where it stands: Mixed, and the geography of the evidence is the story. Against, at this site: limited added value versus calcium alginate on the equine distal limb claim 20unconfirmed; isolated PICO (population, intervention, comparator, outcome)-system NPWT insufficient to stimulate healing in a 2026 equine commentary claim 39unconfirmed. The canine negative pilot is n=5 on hemithorax, not distal limb claim 9unconfirmed. For: fewer complications and less sepsis in 50 dogs claim 34unconfirmed; faster closure in 26 dogs, 14.2 versus 28.6 days claim 120unconfirmed; the human single-use meta-analysis — on closed surgical incisions claim 33unconfirmed. Background: only 36% of equine second-intention experimental studies found a significant improvement claim 121unconfirmed. The format-fit question is sharper than the efficacy one. The cheapest canister-free system in a 2023 comparison is stated in the same sentence to be limited to low-exudate wounds such as closed incisions claim 109unconfirmed, and this candidate targets open and contaminated wounds — which is why dressing capacity is now its own condition claim 152unverified. The one complication in the 7-dog graft series was fluid accumulating in the evacuation tubing claim 35unconfirmed. Which device the unfavourable equine study used is open: its abstract names none claim 73unconfirmed; one review table says V.A.C. ATS with transcription errors claim 74unconfirmed; another says data not presented claim 76unconfirmed. The V.A.C. ATS and its canisters are real cleared products claim 71verified, claim 72verified, but the file refuses to let that stand as a finding about the study claim 62unconfirmed.
What would settle it: A human reads the equine study's Methods for the device, then the 50-dog and 26-dog full texts; then the controlled study the file specifies, at least 20 dogs and/or 20 horses over a minimum 4-week follow-up claim 47unconfirmed.
If it's false: The engineering is solved and there is still no product, because the therapy does not beat a dressing costing a fraction as much.
Would have to be true: Nothing external would force a test battery — so the build plan would have to impose its own, including a home-use usability validation that no regulator will ask for, and a label that may go where the human predicate's does not.
Where it stands: The premarket inversion is verified from primary text: the veterinary registration exemption claim 2verified, the 510(k) (premarket notification) duty conditioned on human use claim 48verified, PMA (premarket approval) scoped to human use claim 49verified. New this build: the human NPWT special controls — usability testing showing users can operate the device from the IFU (instructions for use) alone, prescription use only — sit in a part the regulation scopes to devices intended for human use claim 99verified. So the human counterpart's build spec is Class II (FDA device class II) under code OMP claim 4verified and binds none of this; it is a template to adopt voluntarily. The post-market surface survives: Part 7 "Product" reaches devices intended for animal use claim 84verified, recall is a firm's own removal or correction claim 77verified taken voluntarily claim 78verified; FD&C Act exposure itself is unchecked claim 3unconfirmed, claim 63unconfirmed, claim 133unconfirmed. The home-use label is the live constraint: the Wound Pro Apex IFU is facility-only claim 94verified, and whether a veterinary label may claim post-discharge use is unbacked — the 2026-09-29 attempt crashed on a tooling fault, which says nothing about the IFU claim 141unconfirmed. Sterile-packaging and tooling validation with no reviewer is unchecked claim 31unconfirmed.
What would settle it: Regulatory counsel on what an animal-only label may claim about home use; the OMP special-controls battery, including the IFU-only usability test, written into the design plan as an adopted requirement.
If it's false: The risk is not a blocked pathway but an unforced one: nobody asks for the bench data or the usability test, and the device ships on a seal and an owner workflow that were never characterised.
Would have to be true: A fur-compatible adhesive drape would have to seal over a conical, moving distal limb for days without irritation that breaks the seal — without the adjunctive adhesives already tried and failed — and still coexist with a chew- and bandage-resistant interface in one assembly.
Where it stands: The one equine trial of a canister-free single-use device did not reach its endpoint: the horses readily tolerated it, and difficulties with adhesion and seal stopped the study despite many adjunctive adhesives claim 75unconfirmed. In dogs, a complete seal needed adhesive spray, vacuum maintenance lessened over time, and skin irritation hampered negative pressure in all dogs claim 9unconfirmed. The candidate's own bench conditions are unbacked and unreachable by any connector because the hardware does not exist: the seal claim 28unconfirmed, the interface surviving the animal and the bandage claim 29unconfirmed, irritation below the threshold that breaks the seal claim 30unconfirmed, and the two components combined in one wear assembly claim 51unconfirmed. Whether Kamus's failure is the drape (fixable) or the architecture (not) is on record and unanswered claim 140unconfirmed.
What would settle it: Seal-retention testing on model or cadaveric canine and equine distal limbs through a motion cycle at the target pressure, leak logged; adhesive irritation testing on furred skin; a human reading the Kamus full text for which adhesives failed, where and how. Then the file's own seal study: at least 15 dogs and 15 horses, no more than one unscheduled dressing failure per 72-hour wear interval over a minimum 5-day course claim 46unconfirmed.
If it's false: There is no product. The moat claim 16unconfirmed and the reason to buy this over the human unit claim 19unconfirmed both reduce to this property.
Would have to be true: Seal retention, leak-driven duty cycle, pressure at the dressing port under motion and under fault, exudate capacity, and adhesive irritation would all have to be characterised on a model before any animal is enrolled — and the rig would first have to reproduce the published seal failure.
Where it stands: None of it has been measured claim 28unconfirmed, claim 30unconfirmed, claim 65unconfirmed, claim 66unconfirmed, claim 152unverified. The capital plan goes to prospective case series claim 18unconfirmed — the eventual step, and a costly first one, since the documented failure mode is mechanical claim 75unconfirmed, claim 9unconfirmed. The single-fault pressure limit is specified and untested claim 66unconfirmed.
What would settle it: A limb-model rig logging pressure at the dressing port and pump ON-time against a calibrated leak, with fault injection and the firmware disabled; run first against a commercial PICO dressing to reproduce the Kamus failure, then against the candidate's drape. Tier 3 sits immediately downstream of this slide: no agent here greenlights an animal study.
If it's false: A case series is spent finding a seal problem a rig would have shown first, and the profession reads the result as the therapy failing. If the pressure limit is the thing left uncharacterised, a firmware fault over-pressures a grafted limb on a patient that cannot report it.
Would have to be true: A sterile single-use consumable with a battery pump would have to be manufacturable with margin, second-sourced, shipped legally, applied by ordinary staff, survive an animal and a bandage, and be disposable at the end by a practice or an owner.
Where it stands: Tooling and packaging validation claim 31unconfirmed and margin claim 27unconfirmed are unchecked; supply rests on one adhesive and one converter with no second source named claim 57unconfirmed. Application by ordinary staff is the desirability deck's condition claim 23unconfirmed. The battery path: UN 38.3 before transport matched the regulation text for lithium, while the FCC half is contested — SDoC needs no third-party lab and 15.103 carries exemptions a human must rule on claim 67unconfirmed; IEC 62133-2 versus IEC 60086-4 for primary cells claim 68unconfirmed; IEC 60601-1 read as in scope by its own definition of PATIENT, adopted voluntarily claim 69unconfirmed. Disposal of a lithium-cell pump: the federal universal-waste and household-waste routes both exist in text, and whether the cell is hazardous or what any state requires is undetermined claim 142unconfirmed. The pump-to-dressing joint is where the human category's field failures cluster claim 59unconfirmed, claim 37unconfirmed, alongside a pump failing without alerting the user claim 95unconfirmed, a battery-missing error claim 97unconfirmed and a canister O-ring that would not mate claim 98unconfirmed.
What would settle it: A contract-manufacturing quote against a frozen dressing spec; a second-source plan for the adhesive; the published scope clauses of the two IEC cell standards and of 60601-1; regulatory counsel on 15.103; a waste determination for the cell chemistry chosen.
If it's false: Unit economics and user safety fail at the same place the seal does — in the hands of someone the company never meets.
Would have to be true: The live canister-free absorbent-dressing claims would have to leave a practicable design space for a fur-compatible drape, a limb-conforming interface and a pump-to-dressing connector — so that the workaround is the design, not a compromise of it.
Where it stands: The foundational layer is open: both Wake Forest NPWT patents are expired claim 5verified, claim 6verified. The live layer is named but unread. The claim-1 recitations the file records — a layered absorbent dressing with a filtered port claim 85unconfirmed, canister/canisterless switch detection at the connector port claim 86unconfirmed, a backing layer with adhesive over an absorbent layer and no species limitation claim 87unconfirmed — are all unreached by any connector. Two further patents point at different design choices: a superabsorbent dressing made for remote fluid collection, not canister-free claim 88unconfirmed, and a diameter-adjustable connector claim 90unconfirmed. An expired animal wound wrap that does without adhesive entirely is on record, with only its expiry reached claim 89unconfirmed. The claim-charts are not run claim 64unconfirmed, claim 138unconfirmed; the landscape itself is unsearched claim 7unconfirmed. The patent as a kill and as a moat is argued on the viability deck; here it is a constraint on what the drape, interface and connector may look like.
What would settle it: Counsel charts the five named patents' independent claims against the drape, interface and connector as designed; the design team reads the expired wrap as a possible adhesive-free fallback.
If it's false: The design space collapses onto the one component that is both the only moat claim 16unconfirmed and the likeliest infringement surface — and on this candidate IP (intellectual property) is the only barrier of any kind claim 2verified, claim 17unconfirmed.
Every condition above with nothing verified behind it — five of seven. Slides 3 and 7 carry verified backing on one half (the premarket inversion; the expired foundational patents) and none on the other (the home-use label; the live family). Read this slide first.
The regulatory questions are the settled ones claim 2verified, claim 48verified, claim 49verified, claim 99verified; everything genuinely unknown is engineering and evidence, and the count of verified engineering facts in the file is still zero.
If only one thing from this chair could be checked: whether a fur-compatible adhesive drape can hold a seal at approximately -80 mmHg on a canine or equine distal limb for at least 72 hours of normal ambulation claim 28unconfirmed.
Why this one rather than exudate capacity or the pump: the seal is the failure already observed in this form factor, in these species — the one equine trial stopped on adhesion and seal claim 75unconfirmed — while exudate capacity and pump duty are predicted failures, and the pump's duty budget is set for a dressing held on an ambulatory limb, so it inherits whatever the seal does claim 65unconfirmed. Everything the candidate claims as its own reduces to this property. It is answerable on a bench, with no animal and no regulator, and no connector in this repo will ever answer it.
Naming it is not a recommendation, a gate, or a kill.
_shared/plain-language.md