Reinforcement Learning for Layout Planning – Modelling the Layout Problem as MDP

Unger, Hendrik; Börner, Frank

doi:10.1007/978-3-030-85906-0_52

Cited by 7 publications

(2 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Lastly, [55] provide a formulation of FLPs as an MDP and use OpenAI Gym for their implementation. Their state space representation makes use of stacked RGB visualizations encoding different bits of planning information making it suitable for the use of CNN-based methods.…”

Section: The Openai Gym Toolkitmentioning

confidence: 99%

“…A recent study has shown that ML methods, and RL specifically, has not yet found its way into FLP research [9]. However, recent work [55,27] demonstrates the need for and interest in the application of RL to FLPs. One key challenge to be addressed for the aforementioned goals is to design the problem representations and underlying Markov Decision Processes (MDP) in a way that they can accommodate different sorts of problem types and become insensitive to problem sizes.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

gym-flp: A Python Package for Training Reinforcement Learning Algorithms on Facility Layout Problems

Heinbach

Burggräf

Wagner

2023

Preprint

View full text Add to dashboard Cite

Reinforcement learning (RL) algorithms have proven to be useful tools for combinatorial optimisation. However, they are still underrepresented in facility layout problems (FLPs). At the same time, RL research relies on standardised benchmarks such as the Arcade Learning Environment. To address these issues, we present an open-source Python package (gym-flp) that utilises the OpenAI Gym toolkit, specifically designed for developing and comparing RL algorithms. The package offers one discrete and three continuous problem representation environments with customisable state and action spaces. In addition, the package supports 138 discrete and 61 continuous problems commonly used in FLP literature. The user can choose between numerical and visual output of observations, depending on the RL approach being used. The package aims to facilitate experimentation with different algorithms in a reproducible manner and advance RL use in factory planning.

show abstract

Section: The Openai Gym Toolkitmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

gym-flp: A Python Package for Training Reinforcement Learning Algorithms on Facility Layout Problems

Heinbach

Burggräf

Wagner

2023

Preprint

View full text Add to dashboard Cite

show abstract

Reinforcement Learning for Layout Planning – Automated Pathway Generation for Arbitrary Factory Layouts

Unger,

Börner,

Fischer

2023

Lecture Notes in Mechanical Engineering

View full text Add to dashboard Cite

Multi agent reinforcement learning for online layout planning and scheduling in flexible assembly systems

Kaven,

Huke,

Göppert

et al. 2024

J Intell Manuf

View full text Add to dashboard Cite

Manufacturing systems are undergoing systematic change facing the trade-off between the customer's needs and the economic and ecological pressure. Especially assembly systems must be more flexible due to many product generations or unpredictable material and demand fluctuations. As a solution line-less mobile assembly systems implement flexible job routes through movable multi-purpose resources and flexible transportation systems. Moreover, a completely reactive rearrangeable layout with mobile resources enables reconfigurations without interrupting production. A scheduling that can handle the complexity of dynamic events is necessary to plan job routes and control transportation in such an assembly system. Conventional approaches for this control task require exponentially rising computational capacities with increasing problem sizes. Therefore, the contribution of this work is an algorithm to dynamically solve the integrated problem of layout optimization and scheduling in line-less mobile assembly systems. The proposed multi agent deep reinforcement learning algorithm uses proximal policy optimization and consists of a decoder and encoder, allowing for various-sized system state descriptions. A simulation study shows that the proposed algorithm performs better in 78% of the scenarios compared to a random agent regarding the makespan optimization objective. This allows for adaptive optimization of line-less mobile assembly systems that can face global challenges.

show abstract

Reinforcement Learning for Layout Planning – Modelling the Layout Problem as MDP

Cited by 7 publications

References 17 publications

gym-flp: A Python Package for Training Reinforcement Learning Algorithms on Facility Layout Problems

gym-flp: A Python Package for Training Reinforcement Learning Algorithms on Facility Layout Problems

Reinforcement Learning for Layout Planning – Automated Pathway Generation for Arbitrary Factory Layouts

Multi agent reinforcement learning for online layout planning and scheduling in flexible assembly systems

Contact Info

Product

Resources

About