D4.1 - Taxonomy of optimisation and simulation solutions for Manufacturing and Logistics Processes


 Taxonomy of optimisation and simulation solutions for Manufacturing and Logistics Processes



Executive Summary

In this deliverable a review of current state of the art of optimisation algorithms and simulation procedures, used to solve manufacturing processes in the collaborative context, is presented.

The manufacturing processes are classified using the Supply Chain Operations Reference (SCOR) model views, regarding the sourcing, production, and delivery requirements. For its homogenisation the manufacturing processes identified are mainly classified as Source (S), Make (M) and Deliver (D) Plans. The Plans are organised in sections, following this classification: Plan Source (S) Processes, Plan Make (M) Processes, Plan Deliver (D) Processes, Plan Source & Make (SM) Processes, Plan Make & Deliver (MD) Processes, and Plan Source & Make & Deliver (SMD) Processes. Mixed classifications are due to the characteristics of some plans, which address combinations of S, M or D processes.

The identification of Literature Plans and Algorithms, classified according to the Type of Plans (S, M, D, SM, MD, SMD), has been supported by the analysis of Pilots Plans. Around half thousand potential plans are identified in the literature, among which 101 have been thoroughly analysed. The input received from the Pilots corresponds to a fraction of the 101 of the selected and analysed plans. The plans selected from the literature are characterised by being realistic, implementable, solvable, and computationally efficient. Most of the analysed Literature Plans have been applied and are based on real problems.

A detailed analysis has been performed for each of the Literature Plans regarding the modelling approach, the solution approach, the planning horizon and period, the collaboration level in which the Plan is modelled and the algorithm proposed to solve the Plan; identifying the input data, objectives, and output data associated to the algorithm. Thus, a set of optimisation, heuristic, metaheuristic and matheuristic algorithms are identified, classified and analysed.

From the analysis of Literature Plans and the Industrial Pilot Plans a set of input data, objectives and output data have been identified. In order to homogenise these input data a collection of categories has been created, which will allow unifying all the terms in a consistent way.

As a result of this deliverable, and after the deep analysis performed, it is expected to propose a preliminary selection of Plans and Algorithms candidates to be implemented in D4.2, taking into account the Pilots’ requirements.

The reviewed algorithms, approaches and procedures could be considered when further designing and developing optimisation algorithms, to collaboratively and automatically solve Replenishment (Source, S), Production (Make, M), and Delivery (D) Plans, among the enterprises belonging to a network.

In summary, D4.1 offers an extended review that will serve as a base from which the design and development of algorithms can be built upon.