Status changed from status: Open to status: Closed.

Q1 received a substantive new answer clarifying confidence thresholds and how to handle low-confidence results—mission commanders should configure minimum confidence levels, tools should avoid flooding with noise, and performers should research optimal approaches. Q2 answer was significantly expanded with concrete details: Phase II data will include CAP missions (2 flights, 25-35 min each) in CUI-marked non-compliant A-GRA format; schema documentation uncertain due to data rights; 30-90 day facility access lead time; MD interface implementation varies by platform, so designs must leverage all available data sources.

Q&A Changes Summary

New Question Added:

• Q1 (New): Clarifies uncertainty characterization requirements—whether the system must always return ranked courses of action with confidence values, or can withhold recommendations when data is insufficient.

Questions Renumbered:

• Previous Q1-Q4 are now Q2-Q5 (no content changes to these answers).

Key Change: The addition of Q1 addresses a critical requirement gap around recommendation confidence and data sufficiency thresholds for Phase I performers designing the re-planning capability.

Q&A Changes Summary

New/Updated Answers:

• A4 (expanded): Clarifies Phase I should produce human-readable, explainable recommendations (not automated integration). Mission planning environment integration is "still in-work"; successful designs should provide easily transitionable output formats. Emphasizes this is a new research area seeking original concepts. CUI data handling acceptable in Phase I; facility sponsorship via DD254 at contract award.

• A5 (updated): Confirms Phase I should handle CUI data, but allows use of preferred unclassified alternatives if migratable to standalone government systems for Phase II.

• A6 (new answer): Evaluation criteria: feasibility for Fleet deployment and scalability across heterogeneous platforms. Government seeks performer-defined success metrics; example problem: analyzing two-of-four platform losses in carrier defense and adaptive improvements. Emphasizes rigor in measurement approach.

Key Clarifications:

• No classified data in Phase I
• Explainability and human understanding are critical success factors
• Output format flexibility prioritized over fixed standards compliance
• Team should propose own evaluation methodology

Q&A Changes Summary

New Questions Added (Q1-Q3):

• Q1: Hardware requirements for carrier/airborne hosting, computing resources availability, post-mission-to-replanning timeline (objective vs. threshold), and MIL-STD-882E safety compliance requirements
• Q2: Intended end users, secondary data consumers, and performer access to end users for needs assessment
• Q3: Mission cycle scope (debrief alone vs. full cycle), availability of machine-readable decision records, platform scope (unmanned vs. manned), and access to manned-aircraft mission data

Key Clarifications: These new questions substantially expand requirements definition around operational integration (where in the mission cycle), hardware/infrastructure constraints (carrier vs. airborne computing), safety standards (MIL-STD-882E), and data availability for both unmanned and manned platforms—previously left largely unspecified.

Previous Q&As Renumbered: Original Q1–Q7 are now Q4–Q10 with no answer changes.

Q&A Changes Summary

New Answer Added to Q5: All three emphasis areas (Blue Force optimization, Red Force modeling, autonomous re-planning) should be addressed, but unequal depth is acceptable as capabilities develop at different rates. A-GRA/AMS-GRA documentation is now accessible via public GitLab repos. Review schedule is performer-determined (minimum reporting required).

New Answer Added to Q6: Clarifies that data normalization is a Phase I research task. Systems must ingest heterogeneous, non-standard data formats. While A-GRA is the baseline, existing datasets already deviate from it; overdependence on standards will cause solution failure.

Expanded Answer to Q7: Specifies AFSIM and/or NGTS as Navy-approved simulation environments (performers must obtain access paperwork). Reiterates non-standard data ingestion requirement. Clarifies that CAMP is not relevant. Emphasizes human-interpretable output with transparent supporting rationale—not classified planning standards integration.

Summary of Q&A Changes

New Questions Added (Q1–Q4): Five entirely new Q&As addressing critical Phase I execution details:

• Integration target for re-planning outputs (existing Navy environment vs. human review)
• Security clearance requirements for Phase I (unclassified work permitted without facility clearance)
• Evaluation metrics and evidence standards for feasibility assessment
• Multi-part question on award quantity, data availability, metrics prioritization, and transition ownership

Restructured Existing Content: Original Q1–Q6 renumbered as Q5–Q10 with identical answers preserved, including the detailed A6/A10 response on A-GRA baseline, LVC demonstration data availability, and Phase I/II deliverable expectations.

Key Clarifications in New Answers:

• Phase I can be performed entirely unclassified using synthetic/performer-generated scenarios; CUI data handling deferred to Phase II
• Government will process facility clearance for Phase II if needed
• Modular framework approach acceptable; independent algorithm development not mandatory
• Commercial AI tools permitted in Phase I; transition to Navy-approved environment for Phase II

Status changed from Pre-Release to Open

Q&A Changes Summary

New questions added: Q1 (3 sub-questions on emphasis area priority, A-GRA documentation release, and review preferences), Q2 (data format structure), Q3 (8 sub-questions on Phase II specifics: simulation environment, CAMP integration, A-GRA accessibility, planning timeline, autonomy scope, weapons recommendations), Q4 (A-GRA specification availability), Q5 (algorithm development scope and modular architecture acceptability).

Key clarifications in new answers:

• Q5: Modular architecture with platform-provided data ingestion, analysis, and COA comparison is acceptable; offerors need not independently develop all optimization/threat-modeling algorithms.
• Q3: Asks critical Phase II requirements previously unanswered (simulation environment identity, PMA-281 integration expectations, autonomy modification limits, weapons-employment scope).
• Q4: Directly addresses availability of unclassified A-GRA materials for Phase I.

Reorganization: Original Q1 content moved to Q6 (same answer preserved).

Changes to Q&A:

Q1 received detailed answers addressing: (1) A-GRA as baseline data format; Phase I performers use own frameworks/scenarios; Government will provide CUI autonomous systems LVC demonstration data for Phase 1/2; (2) Phase II data availability confirmed with non-compliant datasets that systems must ingest; (3) Performers may use available tools in Phase I with transferability requirement to secure systems for Phase II; Phase I minimum deliverable is algorithm selection and documented plan; prototype demonstration desirable.

Q2 remains unanswered.

Added 3 new Q&As on Phase I/II mission data sourcing, Government-provided materials timeline, and commercial AI tool usage during Phase I with transition to Navy-approved systems for Phase II. Original Q1 on Blue/Red Force prioritization moved to Q2.

This Q&A clarifies that applicants have flexibility in prioritizing technical approaches for Phase I, as all three components (Blue Force optimization, Red Force modeling, and autonomous re-planning) are acceptable demonstration focuses without a mandated hierarchy.

New opportunity: Automated Post-Mission De-Brief and Re-Planning for Collaborative Combat Aircraft (CCA) Missions