Robert Valner (University of Tartu), Selma Wanna (University of Texas)
In order to rapidly deploy natural language interfaces for robotics applications, we developed the Universal Meaning Representation Format (UMRF) and TeMoto Action Engine. The proposed UMRF acts as a middle ground for combining existing and cutting edge solutions in the domains of natural language processing and grounding. The open-sourced TeMoto Action Engine offers developers an UMRF based action execution management system to use as an interface for natural language control, via grounding the UMRF data structures to executable code. We demonstrate the feasibility of the developed work by executing robotic tasks on ROS via different parsing systems, including Google Assistant.
Video presented at ROSCon 2019. More details and slides are available at: roscon.ros.org/2019/.