A Role-Based Language for Collaborative Robot Applications.

The recent progress in robotic hard- and software motivates novel, collaborative robot applications, where multiple robots jointly accomplish complex tasks like surveillance or rescue scenarios. Such applications impose two basic challenges: (1) the complexity of specifying collaborative behavior and (2) the need for a flexible and lightweight communication infrastructure for mobile robot teams. To address these challenges, we introduce NaoText, a role-based domain-specific language for specifying collaborative robot applications. It contributes dedicated abstractions to conveniently structure and implement collaborative behavior and thus, addresses the complexity challenge. To evaluate NaoText specifications, we introduce an interpreter architecture that is based on representational state transfer (REST) as a lightweight and flexible infrastructure for communication among robot teams. We exemplify the application of NaoText using an illustrative example of robots collaborating in a soccer game and discuss benefits and challenges for our approach compared to state-of-the-art in robot programming.