Reducing costs, improving quality and being competitive have become key words in the mantra that inspires world industry in the quest for market space. Over the last decade, Brazilian companies have invested heavily in information systems to manage cash flows, purchasing, sales and inventory, as well as in the automation of productive tasks. The new frontier of this process is the automation of production management systems, where there are bottlenecks for which planning and controlling software for the productive chain have not indicated satisfactory solutions yet.
To assist with the solution of these problems, 30 researchers from São Paulo studied over the last three years the mathematical and computational tools which productive sector have in hand. The idea is to help develop software for use in the industries. They studied, for example, new focuses for production control systems, with possible applications in the beverage bottling and furniture industries.
The group is made up of seven lecturers and students for master’s and doctor’s degrees linked to four institutions: the Associated Computing and Applied Mathematics Laboratory of the National Institute of Space Research (Inpe), the São Carlos Institute of Mathematical Sciences of the University of São Paulo (USP), the Production Engineering Department of the Federal University of São Carlos (UFSCar), and the School of Electrical Engineering and Computing of the State University of Campinas (Unicamp), where Professor Paulo Morelato França, the coordinator of the project, works.
“The objective was to promote the interchange of the different knowledge and experience acquired at each institution”, says França. Even though the project does envision the drawing up of systems for practical and commercial use, the researchers went after subjects for their studies in the real problems faced by manufacturing companies.
The research project – which spawned 17 dissertations for master’s degrees, 6 theses for doctorates, and the publication of 32 articles in periodicals and 86 at congresses – was based on the conclusion that the majority of commercial software products is not capable of finding solutions for production planning and control problems that involve crucial aspects of costs and capacities. According to França, “the group worked with mathematical algorithms to attain the greatest possible optimization of productive processes”.
The researchers concentrated on four themes: production planning and programming, problems of industrial cutting and of packaging. In the planning area, one of the questions researched was the dimensioning of production batches. How can a bottling plant, for example, use its equipment in the most appropriate way to reduce costs, considering that there are various kinds of packaging and liquids to be packed? Another important necessity is to achieve the greatest economies of scale possible, meeting demand adequately, without detriment to costs with inventory.
The computational system most used to solve this kind of issue is MRP, which stands for Material Requirements Planning. “This system is strongly criticized by its users, particularly those in command of complex productive chains. In general, the system provides a single solution for a production plan, ignoring restrictions on capacity, and it has difficulties in considering the specific issues in each application”, says França.
One of the most original proposals by the team to reduce these shortcomings was the way they addressed problems with cutting materials. The productive chain of a furniture manufacture was analyzed. In this environment, large sheets of wood are cut to obtain smaller items, which will then be processed and assembled to make up an end product. In the industry, it is common to separate the cutting process from the production planning process. The team looked for solutions to transform these processes into a joint task. By doing so, the sizing of the batches can be better planned, and the cutting of sheets can be carried out carefully, reducing the waste of material.
In the programming of production, one of the problems regards the deviation of time from the conclusion of task on a machine, compared with the delivery date. Most studies on the subject consider, as a measure of performance, the length of the average delay and the penalties incurred by these delays. However, with the adoption of production systems of the just-in-time kind, in which delivery at the right time is valued, tasks that are concluded before the date became a problem, because they swell inventory costs. It is a recent area of research, where methods have not been produced capable of encompassing questions like the time taken to prepare machines to carry out their tasks and the idle time dedicated to maintenance, loading and unloading of raw material.
For packaging, solutions were studied for the storage and transport of production, to reduce costs with logistics. Organizing packaging makes it possible to better use space in containers, trucks and supermarket shelves. All the tasks carried out by the group have the purpose of creating technology for the management of the various stages of the productive chain, in such a way as to allow industry to produce in the fastest, cheapest and most efficient way possible. Achieving progress in this area and knowing how to apply it is what distinguishes the winners and losers today in the global market.
Production Planning and Control in Manufacturing Systems (nº 97/13930-1); Modality Thematic project; Coordinator Paulo Morelato França – Unicamp; Investment R$ 88,000.00 and US$ 93,712.00