Reversible control of robots

Abstract

Programming industrial robots is challenging due to the difficulty of precisely specifying general yet robust operations. As the complexity of these operations increases, so does the likelihood of errors. Certain classes of errors during industrial robot operations can however be addressed using reverse execution, allowing the robot to temporarily back out of an erroneous situation, after which the operation can be automatically retried. Moreover reverse execution permits automatically deriving programs that physically reverse the operations of an industrial robot. This can be useful in industrial assembly, where a disassembly program can be automatically derived from the assembly program. In this case study we investigate robotic assembly from the point of view of reversibility, investigating to what extent program inversion of a robotic assembly sequence for a given product can be considered to derive a robotic disassembly sequence for this same product, and investigating to what extent changing the execution direction at runtime (i.e., backtracking and retrying) using program inversion can be used as an automatic error handling procedure. The programming model used to reversibly control industrial robots is based on an abstract semantics-based model, extended with various features required for reversible control of industrial robots in real-world scenarios, and implemented as a domain-specific programming language.