Dynamic production system identification for smart manufacturing systems

Abstract

This paper presents a methodology, called production system identification, to produce a model of a manufacturing system from logs of the system’s operation. The model produced is intended to aid in making production scheduling decisions. Production system identification is similar to machine-learning methods of process mining in that they both use logs of operations. However, process mining falls short of addressing important requirements; process mining does not (1) account for infrequent exceptional events that may provide insight into system capabilities and reliability, (2) offer means to validate the model relative to an understanding of causes, and (3) updated the model as the situation on the production floor changes. The paper describes a genetic programming (GP) methodology that uses Petri nets, probabilistic neural nets, and a causal model of production system dynamics to address these shortcomings. A coloured Petri net formalism appropriate to GP is developed and used to interpret the log. Interpreted logs provide a relation between Petri net states and exceptional system states that can be learned by means of novel formulation of probabilistic neural nets (PNNs). A generalized stochastic Petri net and the PNNs are used to validate the GP-generated solutions. The methodology is evaluated with an example based on an automotive assembly system.

The Status Quo of Process Mining in the Industrial Sector

“…The production operators decide which machines take part in the assembling. Bhogal and Garg [40] Brzychczy et al [41] Celonis AG [42] Celonis AG [43] Denno et al [44] Abonyi and Dörgő [45] Engel and Bose [46] ER et al [47,48] Fleig et al [49] Fluxicon [22,50] Ho and Lau [51] Ho et al [52] Knoll et al [53] Koosawad et al [54] Lee et al [32,55]; Park et al [33,56] Meincheim et al [57] Nagy et al [58,59] Park et al [10] Paszkiewicz [60] Pospíšil et al [34] QPR Software Plc [61] QPR Software Plc [62] Rozinat et al [30,63] Ruschel et al [31] Saravanan and Rama Sree [64] Ulsan Institute [65]; Son et al […”

“…Computer parts D, C [65,66] Automotive parts D, C [43] Mining machine operation D [41] Semiconductors D [67] Electronic chips D, C [36][37][38] Oil refinement D [45] Automotive assembly D [44] Coils D [58,59] Fig. 1.…”

“…Besides, Eissmann can assure quality in labor-intensive process steps by revealing bottlenecks with discovery, e.g., when working with leather whose processing needs to be finished after a fixed amount of time. Denno et al [44] apply process discovery based on genetic programming to incorporate probabilistic and causal information into the discovered model. In their evaluation, the discovered model is used for scheduling.…”

Data-Driven Causal Modeling of the Manufacturing System

“…While a large number of techniques for accomplishing the second stage are available in literature, e.g. [1]- [5], the first stage of the model construction has been also addressed by some researchers. For example, [6] suggests a meta-heuristic method for simultaneous identification of the model structure and associated parameters for linear systems, which is achieved via a constrained multidimensional particle swarm optimization mechanism.…”

Energy-Aware Flowshop Scheduling: A Case for AI-Driven Sustainable Manufacturing

“…The evolution of data-driven system state analysis is encouraging new numerical and logical methods but at the same time implementable. Together, these generate and leverage the concept of 'smart factories' to comprise the next industrial revolution for manufacturing, characterized by increased flexibility, productivity, efficiency, and sustainability, ultimately ensuring competitiveness in the global market [4,5].…”