Lot-splitting decisions and learning effects

We consider a lot-splitting model where the unit manufacturing time follows a learning curve. Our objective is to maximize the net present value of the total revenue collected at the delivery of the sublots. An algorithm is developed to solve the problem. Computational experiments are conducted to study the performance of two ‘convenient’ operational approaches, namely the equal sublot size approach and the equal time interval approach. The computational results suggest that these two solution approaches are nearly optimal and that the net present value of the total revenue becomes less sensitive to the sublot sizes as the learning effect increases.