Capabilities

· Accurate and complete representation of the relevant constraints and optimization objectives … To plan coil cutting lines efficiently one has to consider several dimensions: order requirements like quantities and width, coils characteristics like weights and compliant qualities, machine characteristics like number of cuttings and max width. All these characteristics translate eventually into a significant number of heterogeneous constraints and conflicting objectives that CCO has to represent and manage consistently in order to generate the best solution among all the feasible ones. Let’s just have a slightly more detailed flavor of the constraints and objectives that CCO takes into consideration. On the order dimension: ordered quantity, tolerances (two different thresholds can be handled), residual quantity, width/length and possible alternatives (strips , blanks), min/max diameter of a single strip, min/max weight of a single strip. Concerning coils: width, weight, possible defects, borders, quality, alternative-qualities, internal test rank, internal and external diameter. Concerning machines: inlet / outlet diameters, thickness, max weight, min/max width, number of cuttings (depending on thickness), max width for blanks cutting. Every solution that CCO generates will be optimized and also described in terms of these characteristics.

· Interactive planning workflow… The workflows defining the planning session are very interactive in nature. The operator can actually decide how much he/she wants to drive the system by means of a friendly interface designed for this purpose. In fact, on one extreme he/she could just decide to set only the main parameters like max trim-loss accepted allowing the system optimizing all the rest based on the modeled/customized criteria that capture the company best practice. However, there are cases where he / she might need to drive the system choices more firmly perhaps identifying orders that have to be absolutely included, shrinking the set of coils to be used, forcing the system to allocate the plan on a certain machine. Thanks to these kind of interactive and flexible workflows it becomes quite natural to test deeply and quickly several hypothesis that could not have been even roughly investigated without such a helpful support. A quite simple but common example is the following: a set of orders for a given customer, product, thickness, quality has to be urgently delivered but they do not combine easily in terms of width / weight and it is also quite difficult to find a good plan even taking into consideration other short term orders. To explore the solution space the operator could easily ask CCO to quickly test different scenarios like: what if I enlarge the candidate orders moving the horizon or including open orders? What is the minimum trim loss I have to pay should I consider only short term orders and accept to increase trim loss thresholds ? The different solutions emerging from different scenarios can be easily compared both through synthetic KPIs as well as through more analytical descriptions. At the end, before validating the best solution, the operator can still change manually a few things on the plan while the system will always control the consistency of these changes (should not violate any constraints). Each interesting solution can either be saved or even frozen to allocate material and eventually sent back to Legacy or ERP for plan execution.

· A modular architecture … COIL CUT OPTIMIZER can work, if necessary, with different work stations either dedicated to different product families or even dedicated to different functions (eg : planning, supervisor and cost control, execution control). The technical integration with the ERP or legacy system can be done in three different ways depending on the characteristics of the ERP / legacy architecture (flat files, CCO native DB, CCO dedicated DB tables defined on the company repository or warehouse). All these different methodologies have the same objective : adapt CCO architecture to the already existing information system in such a way that integration effort is minimized and data redundancy is avoided.

BACK