In this talk, I present a multi-objective algorithm to solve the MMAL resequencing problem considering all these aspects simultaneously. I also present empirical results obtained from recorded event data of the production process over 4 weeks after the deployment of our algorithm in a plant. We achieved an improvement of the average
batch size of about 30%, which translates in a reduction of the color changeovers of more than 20%. Moreover, we reduced by 10% the spread factor of cars planned for a specific date, reducing the risk of delays in the delivery. I discuss the effectiveness and robustness of our algorithm in improving production performance and quality, as well as
the trade-offs and limitations involved. Finally, I also discuss the framework from a simplified theoretical point of view.