Simulation-Based Self Awareness at the Edge for Trusted Satellite Autonomy

Over the past decade, the surge in low?Earth orbit constellations amid a congested and contested space environment has underscored the critical need for onboard autonomy. Reliance on ground-based operations creates vulnerabilities, especially when threats evolve faster than human or network responses, potentially neutralizing US advantages in space. In response, Sedaro Corporation is developing a trusted satellite autonomy solution that leverages its cloud-native Sedaro Platform and extends its capabilities to the edge. The Sedaro Platform, honed over five years of R&D with over $10M in investments from USSF, USAF, NASA, and venture capital, offers collaborative modeling, predictive faster-than-real-time simulation, and digital integration. Its modular architecture supports complex, multi-physics system-of-systems simulations, allowing users to design, validate, and deploy high-fidelity models of space missions. Key functions include GIT-based configuration management, intuitive browser interfaces, and API-driven automation—features that enable rapid development and seamless integration with third-party simulators and hardware. To address the challenges of real-time autonomous decision-making, the proposed solution introduces Edge Deployable Simulators (EDSs) and the Sedaro Autonomy Framework for the Edge (SAFE). SAFE is an open-source flight software framework that integrates EDSs with third-party autonomy tools, creating a “trust layer” that validates AI-based planning via simulation-driven risk analysis. This dual-layer approach that uses techniques such as Monte Carlo analysis and optimization algorithms ensures that any autonomy command is both rapid and safe, minimizing risks like asset loss or orbital debris generation. A core innovation is the Sedaro Compiler module, which transforms high-fidelity cloud-deployable simulators (CDSs) into resource-optimized EDSs tailored for edge computing environments. The Compiler supports resource-aware parallelization, dynamic memory management, and runtime parameterization. By generating highly readable, high-performance Rust code, it enhances simulation reliability, observability, and compatibility across diverse onboard computing architectures. Additionally, integrated cosimulation capabilities allow external autonomy software to interact directly with running simulators, further enhancing decision-making fidelity. Recent on-orbit demonstrations, including a 148-simulation Monte Carlo analysis executed on the LIME spacecraft—with performance benchmarks of 109× to 222× real-time—validate the feasibility and robustness of this approach. In Phase II, Sedaro aims to optimize EDS deployment, evolve the SAFE framework to include native command and control interfaces, and demonstrate long-duration autonomous operations on 3U and 12U spacecraft. This work paves the way for a new era of trusted, efficient satellite autonomy, reducing dependency on ground infrastructure and safeguarding US space superiority.