Picture Exp_7 Picture Exp_9 Picture Exp_8 Picture Exp_10 Picture Exp_11 Picture Exp_13 Picture Exp_12

Plant Simulation: Optimization of Steel Structure Manufacturing

Motivation

The current process of designing a new steel fabrication plant encompasses several steps and involves personnel with different competencies. At the beginning, the commercial crew interacts with the customer, in order to gather the plant requirements. Subsequently, the technical team prepares a first draft plant layout, based both on the customer requirements and on the previous experiences related to very well-known plant layout templates. Based on a typical production mix, a production optimization is run in order to verify the expected plant performance.

Since the needed simulation and optimization tools require powerful hardware, this task has to be done by the technical office and cannot be executed at the customer premises. Therefore, a video of the simulation run is recorded by the technical office and sent to the customer. By watching the video, the customer develops a better understanding of the process and, typically, he wants to apply several changes to the layout or to the machines used. Those changes are sent back to the technical team, where a new simulation and optimization task is executed and a new video is generated. This loop is usually repeated until the required level of maturity of the solution is achieved.

Thus, the existing process is time consuming and inefficient because of many iterations, where only the technical team can run simulations and assess the plant productivity, using dedicated workstations. Each production optimization of a typical plant of medium complexity (composed of 4 machining stations, 2 loading bays, 2 unloading bays and the automatic handling system) requires approximately 8 minutes on a high-end, 8 cores desktop PC, while it requires 30 minutes on a normal laptop. Clearly, a 30 minutes window for each optimization is prohibitive in a negotiation with the customer. Each month, at least 10 requests for early design modifications and simulation are sent to the technical office, to start and carry on the negotiation phase and the mentioned iterations (within the average of 20 new negotiations per year).

Goals of the experiment

The experiment is meant to optimize this process and to enable quicker and faster simulation and optimization even at the customers’ site. This vision requires the implementation of two could-based services to simulate and optimize the production of a complex manufacturing system, composed of several machines and conveyors. The two services are coupled with a client application meant to streamline the access and the steps required to successfully simulate and optimize a production plant (i.e. upload of the simulation model, customization of the layout, selection of the production mix and visualization of the results). The technical objective of this experiment is thus to provide the functionalities of the simulation and optimization tools as cloud-based HPC services, in order to achieve the main business goal to empower a wider range of user (i.e. the commercial crew) with quicker simulation and optimization solutions to be deployed at the customer’s premises.

Technical impact

The implementation of the experiment was successful in reducing the time needed to perform an optimization for a layout of medium complexity from 30 minutes to approximately 3 minutes on portable devices, blowing away the barrier that made impractical the use of such tools during the negotiation phase, at the customer’s premises. Modifications can now be applied showing directly the effects of the changes, streamlining the interaction towards the best configuration. With the achieved implementation of the experiment, several direct economic benefits are expected over a short to mid-term period.

Plant Simulation Automobile Light Design

Economic impact

FICEP benefits from a more efficient proposal phase due to the streamlined interaction between the technical team and the commercial crew, now endowed by quick-everywhere simulation and optimization capabilities (for the average plant afore mentioned, the number of iterations is reduced from a minimum of 6 – where each iteration takes 4 man-days – to 2, quantifiable in 4,800 euros savings, not taking into account the improved quality of the service offered). Taking into account the number of negotiations processes initiated per year, which were estimated before in 20 negotiations, this lead to an estimation of 96,000 euros/year savings. Furthermore, collaboration between different FICEP teams located worldwide is boosted, as the results of different layout simulation are stored in the cloud, further increasing the capability to properly address the customer’s needs.

The cloud-based configuration also allowed TTS to develop a new business (and pricing) model: a monthly 100 euros fee in a pay-per-use model allows to reach a wider number of SMEs having a limited expenditure capacity but a strong necessity of simulation functionalities especially during the machine design phase. These companies would benefit from a usage of the platform purchased as a service on demand. Such SMEs usually operate in niche markets providing speciality high-performing machines in small (also one-of) lots. This will result in an increased number of active customers for TTS, with more than 20 additional machine manufacturers using TTS cloud-based solutions, resulting in 80,000 euros of additional sales over a 3 years time horizon starting from the project conclusion, with the creation of 2 new jobs over the same time period.