OUSD (R&E) MODERNIZATION PRIORITY: Autonomy; Artificial Intelligence/Machine Learning TECHNOLOGY AREA(S): Information Systems; Battlespace The technology within this topic is restricted under the International Traffic in Arms Regulation (ITAR), 22 CFR Parts 120-130, which controls the export and import of defense-related material and services, including export of sensitive technical data, or the Export Administration Regulation (EAR), 15 CFR Parts 730-774, which controls dual use items. Offerors must disclose any proposed use of foreign nationals (FNs), their country(ies) of origin, the type of visa or work permit possessed, and the statement of work (SOW) tasks intended for accomplishment by the FN(s) in accordance with the Announcement. Offerors are advised foreign nationals proposed to perform on this topic may be restricted due to the technical data under US Export Control Laws. Please direct questions to the Air Force SBIR/STTR HelpDesk: usaf.team@afsbirsttr.us. OBJECTIVE: Cooperative weapons can increase effectiveness of the warfighter against a peer adversary while providing increased protection of valuable air assets requiring increased stand-off range. More specifically, this work will enhance the effectiveness of collaborative weapons in multi-day campaigns against integrated air defense systems by delivering analytics, diagnostics, and algorithms that enable rapid reprogramming based on the prior day's battle data. DESCRIPTION: US air superiority is being challenged by the fast-paced technological advances of opponent entities. At the same time, US DoD budgetary constraints limit the possible approaches that can mitigate these opponent advances. To maintain air superiority, while satisfying monetary constraints, one intriguing solution is to overwhelm the enemy through the deployment of teams or swarms of weapons. Using a number of significantly low-cost assets provides an economic advantage versus the deployment of a single highly expensive vehicle, and it flips the cost-exchange ratio of the conflict to favor US forces. Battle data analytics and forensics aims to improve cooperative weapon effectiveness through phase-based learning of multi-day or multi-wave missions. This program's intent is to improve decision-making for next day mission with regards to weapon tactics and selection of algorithms for engagement, purely in software. Updates to the weapon software rapidly hinges upon the ability to analyze the data sent back from the weapons regarding its performance. Hence, this work intends to develop and employ algorithms which analyze prior weapon data from previous missions in order to improve weapon and mission effectiveness for future battles. The improvement will come from updating particular models and parameters for the weapons, as well as, selecting appropriate and effective algorithms in real-time based on the analytic tools that are developed. The learning/analytics challenge can be broken into three broad focus areas: red force learning, blue force learning, and autonomy software based learning. Blue force learning is focused on updating parameters and models for blue weapons (e.g. aero model coefficients, control/guidance gains, seeker models, etc.) while red force learning is focused on updating models and parameters associated with red threats, targets, tactics, and capabilities. Autonomy tactics learning is focused on updating and improving cooperative algorithms, behaviors, and plays of the blue weapon salvos in order to improve mission effectiveness. The results of this work will then inform which data is most beneficial to weapon effectiveness, which can then be used to inform datalink and on-board recording requirements. In tandem with algorithm development, we seek to answer three key questions: What information is most important for communication and logging (at the algorithm/decision level)? How to design mechanisms for effective and rapid updating of parameters/algorithms? How to select algorithms based on whatever data is available at the time and how sparse is the data? PHASE I: During phase I, the performers will determine their methodology to address a particular red, blue, or autonomy tactics analytic challenge. They will select a particular algorithmic approach for data analytics rooted in the appropriate areas (e.g., artificial intelligence or machine learning) for implementation for preliminary results. Extensive literature surveys and prior research highlighting the advantages and limitations of the chosen approach is required. PHASE II: A successful phase II effort will constitute the full development of data analytic tools for the red/blue/autonomy challenges. The performer will implement the approach chosen in phase I within AFSIM or another (AFRL-approved) suitable software environment. Connections between offline tools and real-time swarm-based decisions must be developed. Full comparisons of multi-day collaborative missions using the tools with benchmarks against alternative methods are required. Documentation of the implementation including user manuals, theory manuals, examples, and source code with U.S. government data rights is required. PHASE III DUAL USE APPLICATIONS: Phase III will consist of transitioning the software module proven in phase II to existing code bases employed by the DoD and its prime contractors developing next-generation networked munition concepts. This transition will focus on user support or consulting to effectively deploy the software in a R&D or T&E environment. REFERENCES: Kelleher, John D., Brian Mac Namee, and Aoife D'arcy. Fundamentals of machine learning for predictive data analytics: algorithms, worked examples, and case studies. MIT press, 2020; Rizk, Yara, Mariette Awad, and Edward W. Tunstel. "Cooperative heterogeneous multi-robot systems: A survey." ACM Computing Surveys (CSUR) 52.2 (2019): 1-31. Moubayed, Abdallah, et al. "E-learning: Challenges and research opportunities using machine learning & data analytics." IEEE Access 6 (2018): 39117-39138. Kibria, Mirza Golam, et al. "Big data analytics, machine learning, and artificial intelligence in next-generation wireless networks." IEEE access 6 (2018): 32328-32338. Kashyap, Hirak, et al. "Big data analytics in bioinformatics: A machine learning perspective." arXiv preprint arXiv:1506.05101 (2015). Alighanbari, Mehdi, and Jonathan P. How. "Decentralized task assignment for unmanned aerial vehicles." Proceedings of the 44th IEEE Conference on Decision and Control. IEEE, 2005. KEYWORDS: artificial intelligence; data analytics; machine learning; collaborative weapons; heterogeneous agents
