Dynamic Neural UMVs - Enhanced Autonomy for the U.S. Navy

Technical AbstractProject Title: Dynamic Neural UMVs - Enhanced Autonomy for the U.S. NavyIntroduction:Vivum Computing is spearheading innovation in neuromorphic computing and autonomous systems with a focus on maritimeapplications. Building upon foundational research from a 2019 NASA SBIR grant, our Phase II project is dedicated torevolutionizing UMV autonomy. This initiative aims to overcome the limitations of conventional AI and remotely operatedvehicles by leveraging the symbiotic capabilities of CTRNNs and FPGAs. Our approach promises transformative improvementsin UMV operations, addressing the adaptability, efficiency, and robustness challenges faced in dynamic maritimeenvironments. Objectives:- Develop advanced CTRNN-based autonomy software integrated with FPGA technology to enhance UMV adaptability anddecision-making capabilities.- Demonstrate the technology's operational effectiveness in dynamic, uncertain, and GPS-denied maritime environments.- Validate the technology's scalability, energy efficiency, and compliance with safety and regulatory standards.- Explore commercial and defense applications, ensuring the technology's transition to acquisition and full-scale deployment. Methodology:The project employs a structured approach that includes:- Development of Autonomy Software: Leveraging CTRNNs for dynamic control systems and FPGAs for computationalefficiency, we will create adaptive, energy-efficient autonomy software for UMVs.- Simulation-Based Testing and Refinement: Using Gazebo and UUV Simulator, we will conduct extensive simulation testing toevaluate and refine our software's performance in simulated maritime environments.- In-Water Testing with Small-Scale Models and Drones: Technologies will be validated in real-world conditions using platformssuch as OpenROV, BlueROV2, BlueBoat, and LoCO-AUV- Integration and Optimization: Feedback from testing phases will guide the optimization of our software for operationalefficiency, minimal energy consumption, and compatibility with various UMV platforms.- Safety and Compliance Verification: The autonomy software will be validated against maritime regulations and safetystandards. Expected Outcomes:The project anticipates delivering:- A prototype of advanced autonomy software capable of adaptive control and decision-making for UMVs.- Demonstrated superiority of the CTRNN-FPGA integrated system over conventional AI and remotely operated vehicles inadaptability, efficiency, and operational robustness.- A comprehensive validation of the technology's performance in real-world maritime environments.