Research and Development of Elastic Swarms

A natural next step in the evolution of autonomous unmanned systems is the realization of “elastic” architectures for deploying and managing a fleet of platforms across multiple battalions and simultaneous missions.  In an “elastic swarm”, the lines are blurred between the ownership of available UxS resources, and the units / missions that they are eventually assigned to. The current state-of-the-art in flight tested autonomous swarms includes one operator managing less than ten platforms executing a single mission. True elastic swarming will only be achieved when many operators can independently and asynchronously request and task a subset of a much larger swarm to execute many independent and asynchronous missions. In this Phase II project, Toyon proposes Research and Development of Elastic Swarms (RADES) – a direct extension of our Phase I work and realization of the elastic swarming concept.  In Phase I, Toyon collaborated with subject-matter experts in the Air Force community to develop the elastic swarming concept, performed proof-of-concept development of blockchain technology, developed a modeling and simulation foundation for testing of new autonomy components, and identified a clear path to flight-testing and demonstrating our elastic swarming concept to AFSOC. During Phase II, Toyon will develop an elastic swarming prototype that can be demonstrated in simulation and during a live flight test. Toyon will expand upon its CEDAR autonomy foundation by introducing new components for brokering resources and task requests within a team of fully-autonomous unmanned systems. Our approach will be entirely decentralized – without dependence on a centralized command-center or even a master node within the swarm.  To achieve this Toyon will continue its Phase I work to leverage blockchain technology to ensure secure and reliable decentralized identity management and decision making within the swarm.  Toyon will also extend its ongoing work on ATAK to develop an HMI that uses an LLM speech interpretation mode for voice interaction; enabling task requests to come from multiple independent operators.  Finally, Toyon will develop and validate our prototype solution using the Golden Horde Colosseum (GHC).