Multi-Agent Decision Transformers for Dynamic Dispatching in Material Handling Systems Leveraging Enterprise Big Data

TL;DR

It is demonstrated that Decision Transformers can improve the material handling system’s throughput by a considerable amount when the heuristic originally used in the enterprise data exhibits moderate performance and involves no randomness, and when the performance of the dataset is below a certain threshold.

Abstract

Dynamic dispatching rules that allocate resources to tasks in real-time play a critical role in ensuring efficient operations of many automated material handling systems across industries. Traditionally, the dispatching rules deployed are typically the result of manually crafted heuristics based on domain experts' knowledge. Generating these rules is time-consuming and often sub-optimal. As enterprises increasingly accumulate vast amounts of operational data, there is significant potential to leverage this big data to enhance the performance of automated systems. One promising approach is to use Decision Transformers, which can be trained on existing enterprise data to learn better dynamic dispatching rules for improving system throughput. In this work, we study the application of Decision Transformers as dynamic dispatching policies within an actual multi-agent material handling system and identify scenarios where enterprises can effectively leverage Decision Transformers on existing big data to gain business value. Our empirical results demonstrate that Decision Transformers can improve the material handling system's throughput by a considerable amount when the heuristic originally used in the enterprise data exhibits moderate performance and involves no randomness. When the original heuristic has strong performance, Decision Transformers can still improve the throughput but with a smaller improvement margin. However, when the original heuristics contain an element of randomness or when the performance of the dataset is below a certain threshold, Decision Transformers fail to outperform the original heuristic. These results highlight both the potential and limitations of Decision Transformers as dispatching policies for automated industrial material handling systems.

Figures (6)

• Figure 1: A simplified illustration of the material handling system we used in this case study.
• Figure 2: Overview of our proposed framework of using multiple Decision Transformers as dynamic dispatching policies in a multi-agent setting.
• Figure 3: (a) Distribution of throughput between heuristics of different skill levels and Decision Transformers trained on the data generated by the heuristics. (b) Additional comparison of the throughput distributions of Decision Transformers trained on data generated by a deterministic low-skill heuristic (SLL) and on data generated by a stochastic high-skill online RL policy.
• Figure 4: Distribution of throughput of Decision Transformers of different skill levels conditioned on different specified throughput. 'DT-R', 'DT-L', 'DT-M', 'DT-H' denote models that are trained on the 'Random', 'Low', 'Medium' and 'High' datasets respectively.
• Figure 5: UMAP projection of the unique state visitations generated by the six heuristics of different skill level in (a) the $1^{st}$ and $2^{nd}$ dimension and (b) the $2^{nd}$ and $3^{rd}$ dimension.