Combining a Workload Control system with a due date predictive tool based on neural network

What is it about?

Workload Control (WLC) is a production planning and control system that reduces queues and waiting times in manufacturing job shops. Scientific literature often improves the performances of WLC be using complex job-release mechanisms and sophisticated dispatching rules. This approach, however, makes WLC even more complicated, and thus less appealing to industrial implementation. We propose a complementary approach in this work: we introduce a simple WLC system, and we integrate it with a predictive tool that can accurately forecast delivery dates, based upon the state of the system. Forecasts of delivery dates are made with a neural network, and with regression techniques. Results show minimum delays, controlling WIP levels and number of negotiated due dates.

Featured Image

Photo by Possessed Photography on Unsplash

Photo by Possessed Photography on Unsplash

Why is it important?

This is the main contribution of this work is proposing a forecasting approach that can be very useful in many industrial applications. To test this, we simulated a 6-machines job-shop controlled with WLC and equipped with the forecasting system we propose. This system shows robust and quality results, reporting minimal delays, and also controlling average WIP levels and number of negotiated due dates, if compared with a set classical benchmarks.