A framework and method for analysis of feed-forward industrial and manufacturing lines

Cris Koutsougeras,Ahmad Fayed,Mohammad Saadeh

doi:10.1108/jimse-06-2021-0031

Abstract

PurposeThis modeling facilitates the determination of control responses (or possibly reconfiguration) upon such events and the identification of which segments of the pipeline can continue to function uninterrupted. Based on this modeling, an algorithm is presented to implement the control responses and to establish this determination. In this work, the authors propose using Message Queuing Telemetry Transport (MQTT), which is an integrated method to perform the system-wide control based on message exchanging among local node controllers (agents) and the global controller (broker).Design/methodology/approachComplex manufacturing lines in industrial plants are designed to accomplish an overall task in an incremental mode. This typically consists of a sequence of smaller tasks organized as cascaded processing nodes with local controls, which must be coordinated and aided by a system-wide (global) controller. This work presents a logic modeling technique for such pipelines and a method for using its logic to determine the consequent effects of events where a node halts/fails on the overall operation.FindingsThe method uses a protocol for establishing communication of node events and the algorithm to determine the consequences of node events in order to produce global control directives, which are communicated back to node controllers over MQTT. The algorithm is simulated using a complex manufacturing line with arbitrary events to illustrate the sequence of events and the agents–broker message exchanging.Originality/valueThis approach (MQTT) is a relatively new concept in Cyber-Physical Systems. The proposed example of feed-forward is not new; however, for illustration purposes, it was suggested that a feed-forward be used. Future works will consider practical examples that are at the core of the manufacturing processes.

Highlights

This study concerns the control of complex production lines, referred to as production pipelines, in which discrete products are incrementally assembled or produced in various stages following specific processing sequences
The present work targets the analysis of the consequences of partial halting events in production lines in which discrete objects advance from one processing stage to another. This is useful both in the design phase of production lines and for developing strategies to react to node halts during the operating phase
This study focuses on the production line layout and material follow to optimize the production time

Summary

Introduction

This study concerns the control of complex production lines, referred to as production pipelines, in which discrete products are incrementally assembled or produced in various stages following specific processing sequences. The present work targets the analysis of the consequences of partial halting events (possibly due to faults or maintenance) in production lines in which discrete objects advance from one processing stage to another This is useful both in the design phase of production lines and for developing strategies to react to node halts during the operating phase (of which the algorithm presented here is an example). 3.2 Simulating the algorithm The following model (Figure 7), which depicts a complex manufacturing pipeline, will be used to illustrate the effects of node halting according to the analysis introduced in the previous section. This illustrates that this method of analysis may be used to identify weaknesses of a pipeline design and help with the design of more fault-tolerant pipelines. The feeds from nodes 1 and 2 will be blocked, and since nodes 1 and 2 only feed node 9 and do not have alternative routes, they will both need to Downstream

Example analysis 2

Discussion

Conclusion