A review of the applications of multi-agent reinforcement learning in smart factories
Problem
Smart factories must make large numbers of real-time, optimal decisions for machines and material handling despite dynamic, unpredictable events that cannot all be foreseen.
Why it matters
Real-time handling of uncertainty in manufacturing directly affects productivity, efficiency and automation of Industry 4.0 systems.
What the AI did
The paper reviews multi-agent reinforcement learning as a key AI control approach for realizing efficiency, agility and automation under uncertainty in smart factories.
Results
not stated in source
Impact
The evidence only supports that MARL is surveyed as an applicable control approach, not that a specific factory problem was solved.
Record
| Venue | Frontiers in Robotics and AI |
|---|---|
| Published | 2022 |
| Work type | article |
| Authors | Fouad Bahrpeyma, Dirk Reichelt |
| Citations | 76 (OpenAlex, at import time) |
| References | 97 |
| Model or method | multi-agent reinforcement learning |
| Organization | not stated in source |
| Primary source | https://doi.org/10.3389/frobt.2022.1027340 |
| Open access | gold |
| Keywords | multi-agent reinforcement learning, smart factory, Industry 4.0, real-time control, uncertainty |
Importance rating
significant — thousands to millions affected
Real-time optimal decisions under uncertainty across smart factories is framed as a core Industry 4.0 challenge affecting manufacturing broadly, though no scale is quantified.