Reference: GV/2013/034
Period underway: 01/01/2013 – 01/01/2015
Type: Competitive research projects
Amount: 4,800 €
Financed by: GENERALITAT VALENCIANA (Regional Valencian Government)
Principal Investigator: Gómez Gasquet, Pedro
Participant: Gómez-Gasquet, Pedro; Pérez Perales, David; Franco Pereyra, Rubén Darío; Cortés Fibla, Raúl

Description:

This project addresses a decision-making problem in industrial companies that specifically addresses short-term production system management. The background of this project lie in the works conducted in the research by several team members, and is finally shaped by the proposal put forward while preparing the Principal Investigator’s doctoral thesis [1][2][3]; that of a vision of Production scheduling Problem Processes in Hybrid Flow Workshops.

As a result of previous activities, a Software Platform exists called IPSU-MAS (Integrated Platform for Scheduling under Uncertainty based on a Multi-Agent System). This platform is a prototype capable of automating the decision making that relates to the production scheduling and rescheduling process with uncertainty, where quantitative aspects have been developed in-depth. This platform can integrate more models and incorporate new algorithms.

With today’s crisis, the usual need to increase production system flexibility and to improve this system’s general performance has accelerated. Indeed more and more industrial companies combine, if possible, the manufacture of make-to-stock products with the production of make-to-order units. Several comments on this issue can be made, but essentially for this proposal, it is assumed that there products are already promised other ones whose warehouse entry dates can be amended without steering too far away from the customer service objective. Traditionally, programming has been scheduled by considering restrictions that have not allowed the units promised (to order) or destined to stock to be distinguished in the same lot. This proposal contemplates the need to address a decision-making system which, in production scheduling will facilitate the implementation of the lot streaming approach to efficiently adjust the size of lots of the jobs made in different productive resources, conditioned by mixed make-to-stock/make-to-order production planning. The following are proposed for this purpose:

• In processes terms: establishing a framework based on business processes among the parties involved (management, planning, scheduling and operators) that adjusts to the complexity of the make-to-order (MTO)/make-to-stock (MTS) industrial reality.
• In quantitative terms: designing and implementing different variants of an algorithm that will enable the lot streaming concept to be applied to minimise mean delays in Hybrid Flow Workshops.

When the objective is efficient resources use, having to balance different interests usually occurs. Production programmes will have to be generated in this proposal, based on the balance of interests between level of planning and scheduling. So establishing balanced business processes is proposed, which will be closely linked to the type of problem that needs to be solved. It must be from such knowledge (production system types, methods to solve the problem under study, types of characteristic conflicts, etc.) where a strategy can be adopted to reach agreements through negotiation. The combination of MTO/MTS approaches poses significantly different a problem, where the production system is shared by two different management forms. This was why the option of taking advantage of the Research Team’s experience in modelling and problem solving was considered in order linked to production scheduling to achieve this extension in the project subsidised in 2011 by the Universitat Politècnica de València (PPI_06_10) entitled «Models and Methods with Negotiated Procedures based on Software Agents for Production Scheduling in Hybrid Flow Workshops subject to Uncertainty in mixed MTS-MTO environments and its design methodology (NegoSOL-MAS)».