Generalizing Multi-Objective Search via Objective-Aggregation Functions

Hadar Peer

doi:10.1109/lra.2026.3741789

Generalizing Multi-Objective Search via Objective-Aggregation Functions

2026-10-08

0

PRE

AI

DOI：10.1109/lra.2026.3741789

Original

Original request for help

Save

Abstract

En 中文

Multi-objective search (MOS) is essential in robotics, as real-world robotic systems need to simultaneously balance multiple, often conflicting objectives. Recent works explore complex interactions between objectives, leading to problem formulations that do not allow the usage of out-of-the-box state-of-the-art MOS algorithms. We suggest a generalized problem formulation for MOS on graphs that distinguishes between hidden objectives, tracked only during search, and solution objectives, the objectives of actual interest, recovered from the hidden ones via aggregation functions. We show that our formulation supports the application of standard MOS algorithms, necessitating only to properly extend several core operations to reflect the specific aggregation functions employed. We demonstrate our approach in several diverse robotics planning problems, spanning motion planning for navigation, manipulation, planning for medical systems under obstacle uncertainty, inspection planning, and route planning with different road types, and provide empirical evidence that the extended algorithms outperform their vanilla counterparts applied to the same problems without objective aggregation, with runtime speedups that vary by application and reach several orders of magnitude in the most favorable cases.

Keywords:

Motion and Path Planning

Planning under Uncertainty

Generalizing Multi-Objective Search via Objective-Aggregation Functions

Abstract

Journal

Organization

Cited Papers