Mobile Autonomous Robotic Swarms (MARS)

Lunar Outpost’s Mobile Autonomous Robotic Swarms (MARS) is an algorithmic software layer that acts as a low-drag force multiplier for heterogenous, autonomous swarms to enable operation in extreme environments. MARS leverages a distributed, decentralized architecture allowing each asset to individually determine its optimal action given its knowledge of the objective and the swarm. MARS’s heterogenous algorithms enable its deployment on any robotic asset and can handle a consortium of specialized and generalized assets. The distributed and decentralized nature of MARS makes it robust to communication loss and asset loss as any individual asset will always complete the objective when feasible, while other assets within the swarm take over in instances of losses or failures. The MARS algorithm runs locally on each system, making each asset computationally independent and further increasing resiliency. Given an objective, each MARS asset will individually determine the optimal action it should take accounting for its knowledge of the larger swarm, objective, and the platform’s specific sensing proficiencies. Each platform MARS operates on can retain different sensing, computing, or software capabilities, making the solution truly heterogeneous. As part of a SBIR Phase II, MARS has been successfully demonstrated onboard Lunar Outpost robotic platforms, performing a swarm search of an unmapped area to locate, map, and characterize an unknown signal. This demonstration validated that the MARS software layer could interact with heterogenous hardware and software systems to govern the action of multiple robotic systems at the same time to correctly pursue their objective. This demonstration also proved that acquiring and mapping an unknown signal can be performed far more efficiently and precisely through the deployment of swarm assets than with standalone methods. Robotic swarms offer significant benefits in terms of scalability, redundancy, and distributed sensing. Their decentralized nature ensures robustness, as the failure of individual assets doesn't compromise the system's overall functionality. By distributing sensors across multiple robots, they can cover extensive areas in parallel, enhancing task efficiency and environmental perception. Furthermore, without centralized control, swarms can dynamically adapt to environmental changes, making them highly suitable for operations in unpredictable or changing contexts. Lunar Outpost proposes the on-orbit demonstration of Mobile Autonomous Robotic Swarms (MARS) enabled spaceflight robotic systems for this STRATFI effort. This demonstration will include three Lunar Outpost-produced spaceflight assets integrated with the MARS software suite. This STRATFI effort will focus on the development and maturation of the MARS solution while simultaneously leveraging Lunar Outpost heritage spacecraft design to perform the effort in an efficient and cost-effective manner.