This work is directly applicable to improving the capabilities of current NASA robotic systems such as Valkyrie, SSRMS, and Resource Prospector. TRACLabs has contacts with NASA JSC ER robotics specialists including Scott Askew, Dr. Joshua Mehling, Dr. Kimberly Hambuchen, Dr. Ron Diftler, and Dr. Bill Bleuthmann. TRACLabs also has contacts Within NASA MOD ROBO flight controllers, including Quinn Carelock (DX2 branch chief), Scott Wenger (robotics team lead), Jennie Young (USA), and Tifanie Smart (USA) and can leverage these contacts to guide the development of capabilities and human interfaces that NASA robots require. As future NASA robotics missions are expected to rely heavily on dexterous mobile robots such as Valkyrie, these robots will need sophisticated application development and autonomy software in order to function. Such capable systems will assist humans with tasks such habitat construction or geological excavation or will be required to perform autonomous repair tasks on satellites or Deep Space Gateway. The CRAFTSMAN++ toolkit will allow task developers to program these capabilities while also allowing a human teleoperator to monitor and control these robots to ensure mission success.
TRACLabs is already working with a large-scale automotive customer to integrate CRAFTSMAN into a production cell at one of their factories. This "flexible workcell" was installed in the Fall of 2017 and continues to operate 24/7 without human supervision. TRACLabs is currently in negotiations to reproduce the developed cell at another location within the same plant, as well as to adapt the software to additional tasks. The customer's ultimate goal is to build the "factory of the future" where robots can adapt quickly to part location or shape variation and can work safely with humans without the need for barriers or other expensive safety equipment. While the mentioned customer has chosen TRACLabs to be their primary software developments for this effort, the need for flexible robot software in industrial applications is becoming increasingly obvious. As the demands for robust & capable systems grows, application software will need to rely less on tools based on precomputed motions, and more on sensor-driven, reactive techniques that can reliably perform complex assembly tasks. Our software will greatly facilitate this goal. Based on our success to this point, we expect substantial interest from additional commercial partners in automotive, energy, construction, and service domains to integrate our control software (and any improvements we make to it through further development) and have begun discussions to these ends with a number of potential customers.
More »