Metis Technology Solutions proposes a collaborative sensing and environmental modeling approach for IVA robots such as Astrobee and Robonaut 2. IVA robots are intended to autonomously manage habitats and spacecraft, such as Gateway. To do this effectively, IVA robots must be able to collect, fuse, and share information with other IVA robots, as any human team would do to accomplish a task. Metis proposes an online collaborative process or a service occurring on the spacecraft, where any IVA robot, both free flyer and humanoid robots, can contribute new data to a central server for data fusion. The fused data can then be modeled, analyzed, and shared with the other IVA robots through the same central server, thus allowing IVA robots to manage and navigate independently, and as part of a team. Metis proposes to do this by first enabling Collaborative SLAM for Astrobee, an on-line vision-based mapping process that not only establishes the proposed architecture, but also augments Astrobee with the ability to update SLAM maps without the need for ground station involvement. In addition, fusion and co-registration of other IVA robot sensor data (i.e. CO2 data, acoustics data) with SLAM maps will be investigated as they are key to spacecraft maintenance tasks. The technology developed not only solves an operational limitation for Astrobee, but also fills technical gaps identified by the proposed Game Changing Development (GCD) Integrated System for Autonomous and Adaptive Caretaking (ISAAC) project.
The system can be directly utilized by Astrobee to overcome an operational limitation and demonstrate collaborative sensing and modeling using multiple Astrobee robots or a combination of Astrobee robots and Robonaut 2. It can be applied to any autonomous system, with one or multiple agents, for exploration and management of an environment such as teams of ground-based robots managing an area on the moon or mars. Teams of non-autonomous systems also benefit where data from human operated vehicles with sensor payloads are gathered and utilized.
The proposed system can be applied to any autonomous system, with one or multiple agents, for exploration and management of an environment. Collaborative sensing and mapping can be utilized by mobile agents (human operated or autonomous) for exploring and managing environments where environmental information is captured for the first time and needs rapid distribution.