Decentralized and Asymmetric Multi-Agent Learning in Construction Sites
Multi-agent collaboration involves multiple participants working together in a shared environment to achieve a common goal. These agents share information, divide tasks, and synchronize their actions. Key aspects of multi-agent collaboration include coordination, communication, task allocation, coop...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Open Journal of Vehicular Technology |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10715664/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832582283993808896 |
|---|---|
| author | Yakov Miron Dan Navon Yuval Goldfracht Dotan Di Castro Itzik Klein |
| author_facet | Yakov Miron Dan Navon Yuval Goldfracht Dotan Di Castro Itzik Klein |
| author_sort | Yakov Miron |
| collection | DOAJ |
| description | Multi-agent collaboration involves multiple participants working together in a shared environment to achieve a common goal. These agents share information, divide tasks, and synchronize their actions. Key aspects of multi-agent collaboration include coordination, communication, task allocation, cooperation, adaptation, and decentralization. On construction sites, surface grading is the process of leveling sand piles to increase a specific area's height. There, a bulldozer grades while a dumper allocates sand piles. Our work aims to utilize a multi-agent approach to enable these vehicles to collaborate effectively. To this end, we propose a decentralized and asymmetric multi-agent learning approach for construction sites (DAMALCS). We formulate DAMALCS to reduce expected collisions for operating vehicles. Therefore, we develop two heuristic experts capable of achieving their joint goal optimally, by applying an innovative prioritization method. In this approach, the bulldozer's movements take precedence over the dumper's operations. This enables the dozer to clear the path for the dumper and ensure continuous operation of both vehicles. As heuristics alone are insufficient in real-world scenarios, we utilize them to train AI agents, which proves to be highly effective. We simultaneously train dozer and dumper agents to operate within the same environment, aiming to avoid collisions and optimizing performance in terms of time efficiency and sand volume handling. Our trained agents and heuristics are evaluated in both simulation and real-world lab experiments, testing them under various conditions such as visual noise and localization errors. The results demonstrate that our approach significantly reduces collision rates for these vehicles. |
| format | Article |
| id | doaj-art-0eb3438822a04c58bbd3cf56af751aa6 |
| institution | Kabale University |
| issn | 2644-1330 |
| language | English |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Open Journal of Vehicular Technology |
| spelling | doaj-art-0eb3438822a04c58bbd3cf56af751aa62025-01-30T00:04:16ZengIEEEIEEE Open Journal of Vehicular Technology2644-13302024-01-0151587159910.1109/OJVT.2024.347992710715664Decentralized and Asymmetric Multi-Agent Learning in Construction SitesYakov Miron0https://orcid.org/0000-0002-2582-5098Dan Navon1Yuval Goldfracht2Dotan Di Castro3https://orcid.org/0009-0001-0900-3932Itzik Klein4https://orcid.org/0000-0001-7846-0654Bosch Research, Haifa, IsraelBosch Research, Haifa, IsraelBosch Research, Haifa, IsraelBosch Research, Haifa, IsraelThe Autonomous Navigation and Sensor Fusion Lab, The Hatter Department of Marine Technologies, University of Haifa, Haifa, IsraelMulti-agent collaboration involves multiple participants working together in a shared environment to achieve a common goal. These agents share information, divide tasks, and synchronize their actions. Key aspects of multi-agent collaboration include coordination, communication, task allocation, cooperation, adaptation, and decentralization. On construction sites, surface grading is the process of leveling sand piles to increase a specific area's height. There, a bulldozer grades while a dumper allocates sand piles. Our work aims to utilize a multi-agent approach to enable these vehicles to collaborate effectively. To this end, we propose a decentralized and asymmetric multi-agent learning approach for construction sites (DAMALCS). We formulate DAMALCS to reduce expected collisions for operating vehicles. Therefore, we develop two heuristic experts capable of achieving their joint goal optimally, by applying an innovative prioritization method. In this approach, the bulldozer's movements take precedence over the dumper's operations. This enables the dozer to clear the path for the dumper and ensure continuous operation of both vehicles. As heuristics alone are insufficient in real-world scenarios, we utilize them to train AI agents, which proves to be highly effective. We simultaneously train dozer and dumper agents to operate within the same environment, aiming to avoid collisions and optimizing performance in terms of time efficiency and sand volume handling. Our trained agents and heuristics are evaluated in both simulation and real-world lab experiments, testing them under various conditions such as visual noise and localization errors. The results demonstrate that our approach significantly reduces collision rates for these vehicles.https://ieeexplore.ieee.org/document/10715664/Multi-agentsdeep learningdecentralized decision makingconstruction-sites automationlocalization uncertainties |
| spellingShingle | Yakov Miron Dan Navon Yuval Goldfracht Dotan Di Castro Itzik Klein Decentralized and Asymmetric Multi-Agent Learning in Construction Sites IEEE Open Journal of Vehicular Technology Multi-agents deep learning decentralized decision making construction-sites automation localization uncertainties |
| title | Decentralized and Asymmetric Multi-Agent Learning in Construction Sites |
| title_full | Decentralized and Asymmetric Multi-Agent Learning in Construction Sites |
| title_fullStr | Decentralized and Asymmetric Multi-Agent Learning in Construction Sites |
| title_full_unstemmed | Decentralized and Asymmetric Multi-Agent Learning in Construction Sites |
| title_short | Decentralized and Asymmetric Multi-Agent Learning in Construction Sites |
| title_sort | decentralized and asymmetric multi agent learning in construction sites |
| topic | Multi-agents deep learning decentralized decision making construction-sites automation localization uncertainties |
| url | https://ieeexplore.ieee.org/document/10715664/ |
| work_keys_str_mv | AT yakovmiron decentralizedandasymmetricmultiagentlearninginconstructionsites AT dannavon decentralizedandasymmetricmultiagentlearninginconstructionsites AT yuvalgoldfracht decentralizedandasymmetricmultiagentlearninginconstructionsites AT dotandicastro decentralizedandasymmetricmultiagentlearninginconstructionsites AT itzikklein decentralizedandasymmetricmultiagentlearninginconstructionsites |