# FY26 Release 3 — Topic Fit Scan & Submission Plan
**Prepared:** 2026-06-17 · **Bill plan:** `docs/canonical/22_don_r3_pursuit_build_bill.md` · **Prove:** `GET /api/prove/pursuit-r3` · **Synthesizer:** `npm run synthesizer:r3`
**DoN window:** TPOC questions through **2026-06-23** · submit **2026-06-24 – 2026-07-22** noon EDT  
**AFWERX window:** Pre-release now · submit opens **2026-06-24** · close **2026-07-22**  
**Portals:** [navysbir.com](https://navysbir.com/) · [afwerx.com/sbir](https://www.afwerx.com/sbir) · [DSIP topics-app](https://www.dodsbirsttr.mil/topics-app/)

---

## Operator decision (fill by 2026-06-20)

**Doctrine:** Submit every **GREEN** topic after DSIP scan — no arbitrary cap. Skip **RED** (hardware-only, TS/SCI, materials). See `docs/canonical/10_operator_pursuit_stack.md` for the full agreed queue.

| Tier | Component | Topic # | Title | Reuse base | Go? |
|------|-----------|---------|-------|------------|-----|
| **A** | DoN | DON26BX03-NP002 | C-UAS CSO open topic | DITS + ACES + ESMS | ☐ |
| **A** | DoN | DON26BZ03-NV059 | Zero trust combat systems data/access | NV023 ISRT narrative | ☐ |
| **B** | DoN | _scan 24 Jun_ | Maritime / sensors / logistics themes | NV013 · NV032 · POSEIDON | ☐ |
| **B** | AFWERX | _scan 24 Jun_ | Runtime autonomy / COA / disaster prep | NV008 draft | ☐ |

---

## Scan protocol (2026-06-24 morning)

1. Open DSIP → filter **FY26 Release 3** (DoN BAA + AFWERX 26.BZ R3).
2. For each topic, score **GREEN / AMBER / RED**:
   - **GREEN:** live ATLAS module + prior draft exists → pursue
   - **AMBER:** strong fit, no draft → one-week sprint OK
   - **RED:** hardware-only, materials, medical, TS/SCI — discard
3. Run `node scripts/sbir-pipeline.mjs <topic-id>` for each GREEN topic.
4. Log choices in `docs/pipeline/pipeline.json` (`DON-R3`, `AFWERX-BZ-R3` notes).

---

## Expected theme buckets (from DoN pre-release intel)

| Theme | ATLAS modules | Prior art |
|-------|---------------|-----------|
| **Applied AI / ISRT** | ODIN, adaptive AI, NL tactical, sensor fusion | NV023 submitted R1 |
| **Q-BID / quantum** | PQC module only | **POSITION** — do not force unless topic is software-integration |
| **Contested logistics** | Logistics WFF, convoy, Class I–X, DDIL | NV013, sustainment IOS |
| **Maritime C-UAS / ASW** | POSEIDON, maritime COP | NV005, NV025 adjacent |
| **Fire support training** | Training scenario generator, FIRES engine | NV032 submitted R1 |
| **C-UAS / counter-UAS** | ESMS, alert-intel COA, ACES | DLA NV005, Army C-UAS |

---

## AFWERX R3 — likely fit areas (verify at open)

| Theme | Fit | Notes |
|-------|-----|-------|
| Mission readiness / training | **HIGH** | 21-course system, FiST-like scenarios |
| Disaster preparedness | **HIGH** | EM hub, FEMA/ICS — dual-use |
| Applied AI / autonomy | **HIGH** | AGOS, runtime assurance |
| Advanced materials / nuclear | **LOW** | Skip unless software-only subtopic |
| Game-theoretic COA | **HIGH** | Overlaps DISCORD — pick one vehicle |

---

## TPOC question templates (send before 2026-06-23)

**For Applied AI / ISRT-class topics:**
> ISS operates a production multimodal visual-tactical fusion platform with man-in-the-loop approval gates and tamper-evident audit (TRL 6–7). For Phase I, does the Government prefer (a) integration with an existing program-of-record COP, or (b) standalone feasibility on representative exercise data? Is classified data required for Phase I evaluation?

**For contested logistics:**
> Our platform already tracks Class I–X, convoy circulation, and DDIL store-forward. Should Phase I emphasize (a) predictive shortage analytics, (b) in-transit visibility interop, or (c) sustainment WFF dashboarding for expeditionary HQ?

**For runtime autonomy / COA:**
> ATLAS enforces `NO_UNAUTHORIZED_ENGAGEMENTS` and human release at execution boundaries. Does the topic require formal assurance-case artifacts (e.g., DO-178C-style) in Phase I, or operational demonstration with audit logs?

---

## What we are NOT pursuing in Release 3

- Topics requiring **TS/SCI** or classified lab access ISS lacks
- **Hardware-first** sensors, materials, propulsion, directed-energy
- **Blockchain-only** plays (ARM ITV) unless topic explicitly matches hash-chain + logistics narrative
- **Fourth+ submission** while DISCORD and DIANA drafts are incomplete

---

## After topic lock — agent workflow

1. Pull topic details from DSIP API / portal PDF.
2. Fork nearest DOCX from Release 1 (`docs/SBIR_Draft_*`).
3. Run readiness gate (target ≥78/100, 0 RED flags on Phase I deliverables).
4. Operator certify in DSIP → submit before **2026-07-22 12:00 EDT**.

---

## Release 4 (preview — Jul 22 – Aug 19)

Do not start Release 4 drafts until Release 3 is certified. Same scan protocol on **2026-07-22**.