Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

An Anytime, Scalable and Complete Algorithm for Embedding a Manufacturing Procedure in a Smart Factory

Christopher Leet, Aidan Sciortino, Sven Koenig

TL;DR

The paper tackles the Smart Factory Embedding (SFE) problem, which jointly assigns manufacturing procedures to machines and plans agent transport in a smart factory. It introduces TS-ACES, a complete, scalable solver that first builds a traffic-system based embedding via a MILP over epochs, then converts it to a cell-based transport planner $\mathcal{G}_{TS}$ capable of real-time, cyclic operation. By decoupling traffic from fine-grained grid cells and using an epoch-based MILP with hyperparameter search, TS-ACES achieves complete solution quality and scales to 100+ machines, outperforming prior methods on large industrial scenarios. The approach demonstrates practical impact by enabling high-throughput, flexible production planning in large-scale smart factories, with potential for real-time deployment and adaptable traffic-system layouts.

Abstract

Modern automated factories increasingly run manufacturing procedures using a matrix of programmable machines, such as 3D printers, interconnected by a programmable transport system, such as a fleet of tabletop robots. To embed a manufacturing procedure into a smart factory, an operator must: (a) assign each of its processes to a machine and (b) specify how agents should transport parts between machines. The problem of embedding a manufacturing process into a smart factory is termed the Smart Factory Embedding (SFE) problem. State-of-the-art SFE solvers can only scale to factories containing a couple dozen machines. Modern smart factories, however, may contain hundreds of machines. We fill this hole by introducing the first highly scalable solution to the SFE, TS-ACES, the Traffic System based Anytime Cyclic Embedding Solver. We show that TS-ACES is complete and can scale to SFE instances based on real industrial scenarios with more than a hundred machines.

An Anytime, Scalable and Complete Algorithm for Embedding a Manufacturing Procedure in a Smart Factory

TL;DR

The paper tackles the Smart Factory Embedding (SFE) problem, which jointly assigns manufacturing procedures to machines and plans agent transport in a smart factory. It introduces TS-ACES, a complete, scalable solver that first builds a traffic-system based embedding via a MILP over epochs, then converts it to a cell-based transport planner capable of real-time, cyclic operation. By decoupling traffic from fine-grained grid cells and using an epoch-based MILP with hyperparameter search, TS-ACES achieves complete solution quality and scales to 100+ machines, outperforming prior methods on large industrial scenarios. The approach demonstrates practical impact by enabling high-throughput, flexible production planning in large-scale smart factories, with potential for real-time deployment and adaptable traffic-system layouts.

Abstract

Modern automated factories increasingly run manufacturing procedures using a matrix of programmable machines, such as 3D printers, interconnected by a programmable transport system, such as a fleet of tabletop robots. To embed a manufacturing procedure into a smart factory, an operator must: (a) assign each of its processes to a machine and (b) specify how agents should transport parts between machines. The problem of embedding a manufacturing process into a smart factory is termed the Smart Factory Embedding (SFE) problem. State-of-the-art SFE solvers can only scale to factories containing a couple dozen machines. Modern smart factories, however, may contain hundreds of machines. We fill this hole by introducing the first highly scalable solution to the SFE, TS-ACES, the Traffic System based Anytime Cyclic Embedding Solver. We show that TS-ACES is complete and can scale to SFE instances based on real industrial scenarios with more than a hundred machines.

Paper Structure

This paper contains 15 sections, 2 theorems, 16 equations, 2 figures, 2 tables, 5 algorithms.

Table of Contents

  1. INTRODUCTION
  2. Problem Formulation
  3. Manufacturing Procedure
  4. Smart Factory
  5. Agents
  6. Cell Based Embedding
  7. Smart Factory Embedding Problem
  8. Related Work
  9. Overview
  10. Generating a Traffic System Based Embedding
  11. Hyperparameter Search
  12. Generating a Cell Based Embedding
  13. Analysis
  14. Evaluations
  15. Conclusion

Key Result

Theorem 1

An agent in a queue on one of a junction $J_{i} \in \mathcal{J}{}$'s entry roads at the start of epoch $T{}$ takes at most: timesteps to reach a queue on any one of $J_{i}$'s exit roads. Proof. Let $a_{k}$ be the last agent to enter junction $J_{i}$ on epoch $T{}$. Agent $a_{k}$ has to wait $\sum_{R_{j} \in \textsc{entry}{}(J_{i})} totOutB{}(R_{j}, T{}) - 1$ timesteps for the other agents on $J_{

Figures (2)

  • Figure 1: (a) An example manufacturing procedure. Source and sink process are yellow and blue. (b) An example smart factory and (c) its traffic system.
  • Figure 2: TS-ACES's workflow.

Theorems & Definitions (2)

  • Theorem 1
  • Theorem 2