Variants of independence detection in SAT-based optimal multi-agent path finding

Pavel Surynek; Jiří Švancara; Ariel Felner; Eli Boyarski

doi:10.1007/978-3-319-93581-2_7

Variants of independence detection in SAT-based optimal multi-agent path finding

Pavel Surynek, Jiří Švancara, Ariel Felner, Eli Boyarski

Ben-Gurion University of the Negev, Bar-Ilan University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

In multi-agent path finding (MAPF) on graphs, the task is to find paths for distinguishable agents so that each agent reaches its unique goal vertex from the given start while collisions between agents are forbidden. A cumulative objective function is often minimized in MAPF. The main contribution of this paper consists in integrating independence detection technique (ID) into a compilation-based MAPF solver that translates MAPF instances into propositional satisfiability (SAT). The independence detection technique in search-based solvers tries to decompose a given MAPF instance into instances consisting of small groups of agents with no interaction across groups. After the decomposition phase, small instances are solved independently and the solution of the original instance is combined from individual solutions to small instances. The presented experimental evaluation indicates significant reduction of the size of instances translated to the target SAT formalism and positive impact on the overall performance of the solver.

Original language	American English
Title of host publication	Agents and Artificial Intelligence - 9th International Conference, ICAART 2017, Revised Selected Papers
Editors	Joaquim Filipe, Jaap van den Herik, Ana Paula Rocha
Publisher	Springer Verlag
Pages	116-136
Number of pages	21
ISBN (Print)	9783319935805
DOIs	https://doi.org/10.1007/978-3-319-93581-2_7
State	Published - 1 Jan 2018
Event	9th International Conference on Agents and Artificial Intelligence, ICAART 2017 - Porto, Portugal Duration: 24 Feb 2017 → 26 Feb 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10839 LNAI

Conference

Conference	9th International Conference on Agents and Artificial Intelligence, ICAART 2017
Country/Territory	Portugal
City	Porto
Period	24/02/17 → 26/02/17

Keywords

Cost optimality
Independence detection (ID)
Makespan optimality
Multi-agent path-finding (MAPF)
Path-finding on grids
Propositional satisfiability (SAT)
SAT encodings
Sum-of-costs optimality

All Science Journal Classification (ASJC) codes

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-319-93581-2_7

Cite this

Surynek, P., Švancara, J., Felner, A., & Boyarski, E. (2018). Variants of independence detection in SAT-based optimal multi-agent path finding. In J. Filipe, J. van den Herik, & A. P. Rocha (Eds.), Agents and Artificial Intelligence - 9th International Conference, ICAART 2017, Revised Selected Papers (pp. 116-136). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10839 LNAI). Springer Verlag. https://doi.org/10.1007/978-3-319-93581-2_7

Surynek, Pavel ; Švancara, Jiří ; Felner, Ariel et al. / Variants of independence detection in SAT-based optimal multi-agent path finding. Agents and Artificial Intelligence - 9th International Conference, ICAART 2017, Revised Selected Papers. editor / Joaquim Filipe ; Jaap van den Herik ; Ana Paula Rocha. Springer Verlag, 2018. pp. 116-136 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{6ff1f0bb527c494db5b671c5c622f957,

title = "Variants of independence detection in SAT-based optimal multi-agent path finding",

abstract = "In multi-agent path finding (MAPF) on graphs, the task is to find paths for distinguishable agents so that each agent reaches its unique goal vertex from the given start while collisions between agents are forbidden. A cumulative objective function is often minimized in MAPF. The main contribution of this paper consists in integrating independence detection technique (ID) into a compilation-based MAPF solver that translates MAPF instances into propositional satisfiability (SAT). The independence detection technique in search-based solvers tries to decompose a given MAPF instance into instances consisting of small groups of agents with no interaction across groups. After the decomposition phase, small instances are solved independently and the solution of the original instance is combined from individual solutions to small instances. The presented experimental evaluation indicates significant reduction of the size of instances translated to the target SAT formalism and positive impact on the overall performance of the solver.",

keywords = "Cost optimality, Independence detection (ID), Makespan optimality, Multi-agent path-finding (MAPF), Path-finding on grids, Propositional satisfiability (SAT), SAT encodings, Sum-of-costs optimality",

author = "Pavel Surynek and Ji{\v r}{\'i} {\v S}vancara and Ariel Felner and Eli Boyarski",

note = "Publisher Copyright: {\textcopyright} 2018, Springer International Publishing AG, part of Springer Nature.; 9th International Conference on Agents and Artificial Intelligence, ICAART 2017 ; Conference date: 24-02-2017 Through 26-02-2017",

year = "2018",

month = jan,

day = "1",

doi = "10.1007/978-3-319-93581-2_7",

language = "American English",

isbn = "9783319935805",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "116--136",

editor = "Joaquim Filipe and {van den Herik}, Jaap and Rocha, {Ana Paula}",

booktitle = "Agents and Artificial Intelligence - 9th International Conference, ICAART 2017, Revised Selected Papers",

address = "Germany",

}

Surynek, P, Švancara, J, Felner, A & Boyarski, E 2018, Variants of independence detection in SAT-based optimal multi-agent path finding. in J Filipe, J van den Herik & AP Rocha (eds), Agents and Artificial Intelligence - 9th International Conference, ICAART 2017, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10839 LNAI, Springer Verlag, pp. 116-136, 9th International Conference on Agents and Artificial Intelligence, ICAART 2017, Porto, Portugal, 24/02/17. https://doi.org/10.1007/978-3-319-93581-2_7

Variants of independence detection in SAT-based optimal multi-agent path finding. / Surynek, Pavel; Švancara, Jiří; Felner, Ariel et al.
Agents and Artificial Intelligence - 9th International Conference, ICAART 2017, Revised Selected Papers. ed. / Joaquim Filipe; Jaap van den Herik; Ana Paula Rocha. Springer Verlag, 2018. p. 116-136 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10839 LNAI).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Variants of independence detection in SAT-based optimal multi-agent path finding

AU - Surynek, Pavel

AU - Švancara, Jiří

AU - Felner, Ariel

AU - Boyarski, Eli

PY - 2018/1/1

Y1 - 2018/1/1

N2 - In multi-agent path finding (MAPF) on graphs, the task is to find paths for distinguishable agents so that each agent reaches its unique goal vertex from the given start while collisions between agents are forbidden. A cumulative objective function is often minimized in MAPF. The main contribution of this paper consists in integrating independence detection technique (ID) into a compilation-based MAPF solver that translates MAPF instances into propositional satisfiability (SAT). The independence detection technique in search-based solvers tries to decompose a given MAPF instance into instances consisting of small groups of agents with no interaction across groups. After the decomposition phase, small instances are solved independently and the solution of the original instance is combined from individual solutions to small instances. The presented experimental evaluation indicates significant reduction of the size of instances translated to the target SAT formalism and positive impact on the overall performance of the solver.

AB - In multi-agent path finding (MAPF) on graphs, the task is to find paths for distinguishable agents so that each agent reaches its unique goal vertex from the given start while collisions between agents are forbidden. A cumulative objective function is often minimized in MAPF. The main contribution of this paper consists in integrating independence detection technique (ID) into a compilation-based MAPF solver that translates MAPF instances into propositional satisfiability (SAT). The independence detection technique in search-based solvers tries to decompose a given MAPF instance into instances consisting of small groups of agents with no interaction across groups. After the decomposition phase, small instances are solved independently and the solution of the original instance is combined from individual solutions to small instances. The presented experimental evaluation indicates significant reduction of the size of instances translated to the target SAT formalism and positive impact on the overall performance of the solver.

KW - Cost optimality

KW - Independence detection (ID)

KW - Makespan optimality

KW - Multi-agent path-finding (MAPF)

KW - Path-finding on grids

KW - Propositional satisfiability (SAT)

KW - SAT encodings

KW - Sum-of-costs optimality

UR - http://www.scopus.com/inward/record.url?scp=85049359290&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-93581-2_7

DO - 10.1007/978-3-319-93581-2_7

M3 - Conference contribution

SN - 9783319935805

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 116

EP - 136

BT - Agents and Artificial Intelligence - 9th International Conference, ICAART 2017, Revised Selected Papers

A2 - Filipe, Joaquim

A2 - van den Herik, Jaap

A2 - Rocha, Ana Paula

PB - Springer Verlag

T2 - 9th International Conference on Agents and Artificial Intelligence, ICAART 2017

Y2 - 24 February 2017 through 26 February 2017

ER -

Surynek P, Švancara J, Felner A, Boyarski E. Variants of independence detection in SAT-based optimal multi-agent path finding. In Filipe J, van den Herik J, Rocha AP, editors, Agents and Artificial Intelligence - 9th International Conference, ICAART 2017, Revised Selected Papers. Springer Verlag. 2018. p. 116-136. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-93581-2_7

Variants of independence detection in SAT-based optimal multi-agent path finding

Abstract

Publication series

Conference

Keywords

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Cite this