Specifying and achieving goals in open uncertain robot-manipulation domains

TL;DR

This paper presents an integrated approach for robots to understand and achieve goals in open, uncertain environments by combining formal goal specification with hierarchical planning, state estimation, and execution.

Contribution

It introduces a unified system that interprets formal goal descriptions and robustly executes them in real-world, uncertain robot manipulation scenarios.

Findings

Effective goal interpretation in uncertain environments

Robust execution despite occlusion and sensing errors

Integration of planning, estimation, and execution

Abstract

This paper describes an integrated solution to the problem of describing and interpreting goals for robots in open uncertain domains. Given a formal specification of a desired situation, in which objects are described only by their properties, general-purpose planning and reasoning tools are used to derive appropriate actions for a robot. These goals are carried out through an online combination of hierarchical planning, state-estimation, and execution that operates robustly in real robot domains with substantial occlusion and sensing error.