NAVWAR Open Topic for Unified Assured Positioning, Navigation, and Timing Operational Awareness and Decision Support

# Summary of Q&A Changes **Key Updates:** 1. **A1 (NEW ANSWER):** Phase II will NOT require facility clearance—unclassified work with GPNTS ICDs as Government-furnished information is acceptable. 2. **A2 (NEW ANSWER):** Government explicitly does NOT expect human research/IRB determinations under this SBIR, eliminating compliance burden for Phase I operator evaluations. 3. **Q10 (NEW MULTI-PART QUESTION):** Five detailed technical clarifications added: - **Level of APNT data:** Performers should focus on processed source-level outputs (health, integrity, confidence, degradation indicators) rather than raw measurement data - **Anomaly diagnosis:** Correlation of existing APNT system indicators is sufficient; Phase I doesn't require independent anomaly classification from raw patterns - **Composite confidence:** Transparent rule-based/weighting approaches acceptable; machine learning/statistical fusion algorithms not required for Phase I - **Recovery logic:** Performers may define representative fallback rules and validate with SMEs; Government will provide authoritative procedures in Phase II - **Sub-second latency:** (Answer appears truncated in update) **Impact:** Answers 1 & 2 reduce regulatory/compliance risk. Q10 answers significantly lower Phase I technical bar—rule-based correlation sufficient; statistical/ML confidence fusion not mandatory.

# Q&A Changes Summary **Two new questions added:** - Q1: Phase II clearance/classification requirements (answer pending) - Q2: Human research/IRB requirements for operator evaluation (answer pending) **Five previous "Response Pending" answers now completed:** - Q3: Confirmed commercial platforms are responsive - Q4: Phase I remains unclassified; use open-source ASPN/pntOS synthetic data - Q5: Commercial container runtimes acceptable; Phase I platform-independent - Q6: UX familiarity target for ECDIS (not strict standards compliance) - Q7: Any ASPN/pntOS versions acceptable; GPNTS details in Phase II **Major clarifications in existing answers:** - Q8: Navy GPNTS is primary integration target; no shore-based system interfaces required - Q10: New detailed answer clarifies data levels, anomaly diagnosis, confidence algorithms, decision logic, and latency boundaries—establishing that Phase I may use rules-based approaches with source-provided indicators, but advanced ML is encouraged if explainable; sub-second latency applies to initial alerts only, with detailed rationale available via drill-down.

# Q&A Changes Summary **Five new questions added (Q1–Q5)** clarifying Phase I implementation details: - **Commercial platform adaptation** is acceptable (Q1) - **Phase I remains unclassified**; CUI threat taxonomies need unclassified stand-ins (Q2) - **Commercial container runtime acceptable** for Phase I demo; Iron Bank required for Phase II (Q3) - **ECDIS-like UX is the target**; specific S-57/S-52/NMEA standards deferred to Phase II (Q4) - **Publicly available ASPN/pntOS schemas acceptable**; GPNTS-unique messages unavailable until Phase II (Q5) Original Q1–Q3 renumbered to **Q6–Q8** with identical answers preserved.

# Q&A Summary of Changes **Updated A1:** Added detailed guidance on Phase I approach—performers may define representative rules/TTPs; Government will provide authoritative documentation and SMEs post-award in Phase II. Clarified primary integration target is Navy GPNTS (standalone interface following ECDIS conventions); no shore-based system integration required. Moved technical architecture details (API-first, 3-8 sources at 1-10 Hz) from A4 to A1. **New A2:** Confirmed Phase I analysis should focus primarily on APNT-source data displayed on navigation-style interface; broader mission data not planned for ingestion, but modular design encouraged for future accommodation. **New A3:** Significantly expanded answers to Krishna's detailed questions (previously pending). Key clarifications: - Solution must ingest both processed indicators AND derive independent composite confidence via analytics - Rules-based correlation acceptable for Phase I; advanced statistical/ML encouraged if explainable and edge-compatible - Sub-second latency applies end-to-end (ingestion → alert presentation) - Progressive disclosure model approved: initial alert sub-second; detailed rationale/evidence drill-down may follow **A4 Status:** Answer text appears truncated in updated version but core content retained (awards based on funding/quality; Phase II data provided; Phase I uses synthetic/open-source; air-gapped, 100% local processing required; GPNTS is transition owner).

# Q&A Changes Summary **New Answer Added to Q4:** Clarified that Phase I can use open-source simulated data adhering to standards for UX/HMI demonstration; Phase II will receive actual representative datasets. Specified compute platform independence with API-first design for 3-8 PNT sources at 1-10 Hz. Confirmed metrics prioritize operator effectiveness (speed to decision, threat comprehension, decision accuracy, reduced cognitive workload). Emphasized 100% air-gapped, locally-executed capability requirement with no remote connectivity dependency. **New Answer Added to Q5:** Partially completed—restated Phase II GPNTS ICD provision and dual requirement (visualize existing confidence AND derive composite assessment). Confirmed anomaly triage and fallback execution as priority decisions but answer appears truncated. **Key clarifications:** Architecture must be fully self-contained aboard ships; synthetic data acceptable Phase I; specific ASPN/pntOS versions/ICDs deferred to post-award Phase II.

**New Q&A Added:** Q1 (new): Asks about integrating additional operational data sources (weather, mission planning, threat intelligence, platform health) and whether Government will provide Navy doctrine, TTPs, confidence thresholds, or recovery procedures. Also clarifies priority operational functions and required system interfaces. **Key Change:** Previously unanswered questions (Q2-Q5 in original) now appear as Q3-Q6 with numbering adjusted. The substantive Q&A content remains largely unchanged, suggesting the update primarily reorganized existing questions and added the new Q1 to address broader integration and operational guidance expectations.

# Summary of Q&A Changes **New Questions Added:** - Q1: Independent analysis for situational understanding (e.g., assessing degraded sensor impact) - Q2: 7 detailed questions on Phase I data requirements, anomaly diagnosis responsibility, composite confidence methodology, fallback decision logic, and sub-second latency boundaries - Q3: 7 questions on award quantity, government-furnished data/interfaces, evaluation metrics, use of synthetic data, hardware constraints, and transition ownership **Key Clarifications in New Answers:** - **Phase I Data Level:** Performers should model processed source-level outputs (health, integrity, confidence); raw measurement data interpretation acceptable but not required for feasibility - **Anomaly Classification:** Capability expected to correlate APNT-provided indicators; independent classification from multi-source patterns acceptable for Phase I - **Confidence Fusion:** Transparent analytical approaches (rules, weighting) acceptable for Phase I; ML/statistical algorithms not mandated - **Fallback Logic:** Bounded representative decision rules acceptable; Government will validate with SMEs rather than prescribe authoritative procedures - **Latency Split:** Sub-second applies to initial recommendation; detailed rationale/evidence may be presented via drill-down immediately afterward

Status changed from Pre-Release to Open

# Q&A Changes Summary **New/Updated Answers:** - **A1 (Data Availability)**: Clarified that performers will receive GPNTS ICDs post-award; solution must both visualize source confidence AND derive composite assessment; focus on anomaly triage and fallback execution decisions. - **A2 (7 sub-questions)**: Major clarifications on deployment (100% air-gapped, locally executable), architecture (API-first, 1-10 Hz sources, sub-second latency), evaluation metrics (speed to decision, threat comprehension, decision accuracy, workload), interface approach (can integrate or standalone, must follow ECDIS conventions), and historical replay (desirable for short-term review). - **A3 (Decision Traceability)**: Confirmed auditable history linking condition→evidence→action→outcome is "highly desirable" for Phase I and essential for operator trust. - **A4 (Operational Context)**: Limited Phase I scope to APNT-source data only (displayed on nav display); broader mission context deferred due to data availability constraints. **Key Technical Requirements Added:** - 100% air-gapped, locally-executed capability (no remote cloud/shore connectivity) - Sub-second end-to-end latency required - Navy ECDIS interface conventions mandatory - Modular, API-first architecture

# Summary of Q&A Changes **New questions added:** - Q2 (Krishna/VaultIQ): 7 detailed questions on recommendation explainability, deployment architecture (isolated vs. centralized), GPNTS/ASPN/pntOS interface specifications, real-time performance requirements, Phase II evaluation metrics, ECDIS integration approach, and historical analysis/replay capabilities - Q3: Decision traceability and auditable history linking conditions, data, actions, and outcomes - Q4: Operational context beyond APNT data (mission phase, route, platform condition, priorities) - Q5: Governed action routing with operator approval authority **Key clarifications provided:** - Phase II will require actual Navy data (GPNTS, ASPN/pntOS, alternate-PNT interfaces) with confidence/integrity metadata - System should derive independent confidence assessments and map to mission impact, not merely visualize existing confidence - Deployment must support isolated, ship-based execution in contested environments without reliable remote connectivity - Operators should have access to underlying data, reasoning, and evidence supporting system recommendations - Solution may combine APNT status with broader operational context (mission phase, route, platform condition, priorities) - Recovery actions may be routed for operator review/approval while preserving final decision authority

# Q&A Changes Summary **Q1 received comprehensive answers** clarifying Phase I expectations: - Essential operator interface elements defined (system health, confidence, threats, impacts, recovery actions) - Deep focus on **one representative use case preferred** over multiple platforms - Phase I deliverable: **low to medium fidelity demo** with design concepts and feasibility evidence - **Containerization strongly encouraged** as proof-of-concept during Phase I - Platform-independent, modular architecture with Navy interface mapping plan acceptable for Phase I **Q2 answer began** (text cut off) but started addressing solution scope openness. **Key clarifications:** Modularity/API-first design required; containerized PoC using DoD Platform One Iron Bank recommended; human-centered design metrics (speed to decision, accuracy, decision accuracy, training transfer) expected for evaluation.

This Q&A clarifies that Phase I should use simulated data and operator interviews rather than requiring integration with actual Navy systems, and confirms the Government is open to multiple solution approaches (commercial adaptations or purpose-built software) while leaving specific platform focus and integration details for performer determination.

New opportunity: NAVWAR Open Topic for Unified Assured Positioning, Navigation, and Timing Operational Awareness and Decision Support