PLANCORTE – Cutting Plan Optimization
UniSoma developed, for Wirex Cable, a cable cutting plan optimization system for minimizing losses and maximizing service level.
At Wirex, conductor cable with different lengths are kept in stock to meet, through a cutting process, a portfolio of orders. This variety is due to, in general, previous cuts.
The company’s business analysts had to, on a daily basis, face the arduous routine of generating cutting schedules using the available stock with several different cable lengths to meet the various client demands of different cable lengths. A cutting schedule is composed of a set of interrelated decisions: (a) selection of coils in stock, (b) associating each demand with a specific coil, (c) cut sequencing for each combination of orders and coils and (d) determining the length of each cut (orders allow a tolerance for delivered cable lengths).
The selection of coils in stock critically determines the necessary maneuvering in the cutting courtyard, a typical process bottleneck. This selection, associated to combining orders to coils, determines the quantity of generated scrap, the quantity of restocked cables (there is, for every type of cable, the concept of a scrap point, which determines a minimal cable length after which a cable losses sale value) and the level of service for meeting the demand portfolio.
In early 2001, Wirex Cable hired UniSoma to develop and implement a decision support system to determine the cutting scheduling for cable coils, with the objective of improving the level of service in delivering orders from a demand portfolio and minimize raw material losses (generation of scrap).
There was a necessity to create a tool capable of generating, dynamically, cutting patterns that could meet the constant demand variations and optimize the coil maneuvering.
The adopted methodology should allow for a quick rescheduling of combinations between coils and orders to meet special orders and allow the commercial area to simulate order by means of executing various what-if scenarios.
UniSoma developed the PlanCorte system, an AIMMS© based application, composed of a mathematical model based on Column Generating techniques, classically used in solving cutting problems. PlanCorte is a multi-objective decision support tool that optimizes the cutting schedule to improve the service level for meeting portfolio demands and, simultaneously, minimize scrap generation.
The cutting schedule proposed by PlanCorte is based on the generation of optimized combinations and cutting sequences. The combinations indicate which orders and length, within the established tolerances, should be produced from cables in stock. The cutting sequencing minimize distortions caused by errors and uncertainties in stored information on available stocked cable lengths.
The system includes a scenario manager that allows the user to simulate cutting schedules with demand forecasts or orders under consideration. The generation of scenarios is achieved through the integration to EMS-DATASUL© transactional data base. Results suggested by PlanCorte and validated by the company are published as production orders in EMS.
To facilitate the interaction between users and the mathematical model, a friendly graphical user interface (GUI) was incorporated. This interface allows (a) triggering of imports and exports to EMS-DATASUL©, (b) visualization of all imported information, (c) setting of optimization options, (d) display in the form of tables, graphs and maps, of various aspects of the generated optimized solution and (e) validation and manual modification of PlanCorte’s proposed solution.
The use of PlanCorte in the cable cutting scheduling routine at Wirex allowed the reduction of 40% of scrap generated during the process. Because the system allows the optimized use of each coil, maneuvering was significantly reduced and consequently raised the number of daily cuts by 50% and increased the level of orders met.
“The implementation of the cutting optimization system, allowed us to reduce the total generated scrap in up to 40%. Today, PLANCORTE checks the demand portfolio, in a determined period, and evaluates with the available stock, which coils should be used for cutting; grouping cutting orders of a same type of item, and attempting to fully exhaust each coil at once, considering each order length and its tolerances, drastically reducing scrap generation, and reducing the cutting time for a same number of orders.”
Luiz Carlos Sacchi – Supply Chain Management Coordinator – May 2002
|AIMMS Service Partner UniSoma|