We introduce a novel diagnostic reasoning method for robotic systems with multiple robots, to find the causes of observed discrepancies relevant for plan execution. Our method proposes (i) a systematic modification of the robotic action domain description by utilizing defaults, and (ii) algorithms to compute a smallest set of diagnoses (e.g., broken robots) by means of hypothetical reasoning over the modified formalism. The proposed method is applied over various robotic scenarios in cognitive factories.
CITATION STYLE
Erdem, E., Patoglu, V., & Saribatur, Z. G. (2015). Diagnostic reasoning for robotics using action languages. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9345, pp. 317–331). Springer Verlag. https://doi.org/10.1007/978-3-319-23264-5_27
Mendeley helps you to discover research relevant for your work.