We propose to research, develop, and support open-source whole-body control and walking algorithms as well as diagnostics algorithms for the NASA JSC valkyrie humanoid robot. These algorithms will…

This is a $1,243,530.56 Cooperative Agreement awarded by the National Aeronautics and Space Administration (NASA) to the Florida Institute for Human & Machine Cognition (IHMC), a non-profit research institute. The contract is for research, development, and support of open-source whole-body control and walking algorithms, as well as diagnostics algorithms, for the NASA JSC Valkyrie humanoid robot. IHMC will extract and package key foundational algorithms as software libraries, provide API and ROS support, and make the software open source. The goal is to advance humanoid robotics capabilities and enable their use in future space missions. The contract does not have a set-aside designation.

We propose to research, develop, and support open-source whole-body control and walking algorithms as well as diagnostics algorithms for the NASA JSC valkyrie humanoid robot. These algorithms will be based on existing work we have been doing for our national robotics initiative (NRI) and related humanoid projects. The NRI work is focused on expanding the boundaries of humanoid robot capabilities. This project will be focused on providing the core capabilities underlying that work in stable and reliable libraries. We will extract and package key foundational algorithms as software libraries and provide direct API support as well as ROS support. We will provide these libraries to research groups which have access to the valkyrie platform, either through purchase or through the new NSF/NASA valkyrie build initiative. These research groups can then modify the algorithms as needed. We will provide support to the research groups by remote assistance and site visits as appropriate. The delivered software will be made open source and posted on a public code repository. To ensure a high degree of reliability we will develop an extensive suite of automatic test cases. These tests will range from unit tests of each individual software modules to end-to-end tests that verify the full functionality of simulated robots. In addition, we will create an extensive sequence of manual tests, which will be applied to a real valkyrie robot on a frequent basis. We will also develop a series of diagnostic routines, which will be useful for testing and debugging hardware, and which will also be used as acceptance criterion for the groups receiving valkyrie robots. We believe this effort will help advance humanoid robotics and enable them to add value to future space missions.