Collected molecules will appear here. Add from search or explore.
A discrete event simulation framework using Simpy to evaluate model-based and model-free deep reinforcement learning for manufacturing dispatch and queueing optimization.
Defensibility
stars
21
forks
5
The project is a 7-year-old research artifact with very low adoption (21 stars) and zero current activity. It serves as a proof-of-concept for applying DRL to manufacturing scheduling using the Simpy library. While the niche (industrial engineering + RL) is interesting, this specific repository lacks the software engineering rigor, documentation, or community necessary to be considered a viable tool for production or further development. It is essentially a 'frozen' reference implementation of a paper or thesis project. Modern alternatives like Ray Rllib for the RL component and more robust digital twin platforms or updated Simpy-based frameworks (like Salabim) have long since surpassed the capabilities shown here. Its only value today is as a historical reference for how to bridge DES with early DRL approaches.
TECH STACK
INTEGRATION
reference_implementation
READINESS