Algorithms for a Temporal Decoupling Problem in Multi-Agent Planning

Abstract

The Temporal Decoupling Problem (TDP) arises when a group of agents collaborating on a set of temporally-dependent tasks seek to coordinate their execution of those tasks by applying additional temporal constraints sufficient to ensure that agents working on different tasks may operate independently. This paper: (1)formally defines the TDP, (2) presents theorems that give necessary and sufficient conditions for solutions to the TDP, (3)presents a family of sound and complete algorithms for solving the TDP, and (4) compares the performance of several variations of the basic algorithm. Although this work was motivated by a problem in collaborative multi-agent planning, it represents a contribution to the theory of Simple Temporal Networks that is independent of the motivating application.

Cite

Text

Hunsberger. "Algorithms for a Temporal Decoupling Problem in Multi-Agent Planning." AAAI Conference on Artificial Intelligence, 2002. doi:10.5555/777092.777166

Markdown

[Hunsberger. "Algorithms for a Temporal Decoupling Problem in Multi-Agent Planning." AAAI Conference on Artificial Intelligence, 2002.](https://mlanthology.org/aaai/2002/hunsberger2002aaai-algorithms/) doi:10.5555/777092.777166

BibTeX

@inproceedings{hunsberger2002aaai-algorithms,
  title     = {{Algorithms for a Temporal Decoupling Problem in Multi-Agent Planning}},
  author    = {Hunsberger, Luke},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2002},
  pages     = {468-475},
  doi       = {10.5555/777092.777166},
  url       = {https://mlanthology.org/aaai/2002/hunsberger2002aaai-algorithms/}
}