Handling abort commands for household kitchen robots

TL;DR

This work treats abort commands for household kitchen robots as planning/replanning problems and introduces an Abort Task module built on the AbeSim simulator to autonomously reconfigure actions when cancellations occur. The approach uses a PDDL2.1 domain with derived predicates to model safe, stable states, and leverages DBpedia ontologies to augment scene understanding via SPARQL queries. A mapper translates the real-time simulator state into PDDL problems, enabling the planner to generate fallback sequences that leave the kitchen in a safe condition. Experiments with two scenarios show the system can produce concrete plans to clean up after aborts, though limitations include non-optimal plans, lack of storage-fullness modeling, and reliance on a single planner; future work targets multiple planners and richer knowledge integration to improve robustness and efficiency.

Abstract

We propose a solution for handling abort commands given to robots. The solution is exemplified with a running scenario with household kitchen robots. The robot uses planning to find sequences of actions that must be performed in order to gracefully cancel a previously received command. The Planning Domain Definition Language (PDDL) is used to write a domain to model kitchen activities and behaviours, and this domain is enriched with knowledge from online ontologies and knowledge graphs, like DBPedia. We discuss the results obtained in different scenarios.

Figures (2)

• Figure 1: System architecture
• Figure 2: Fetching a knife in the Abe Sim