Pickup and Delivery with Smart Service Times

The research problem to be addressed is the pickup and delivery problem (last mile) with precedence constraints for the pickup/delivery operations, and time windows. The objective is: (i) the design of a new algorithm that will use the algorithm of Curtois et al. [6] as a starting point, and to generalize it with precedence constraints, time windows and smart service times constraints and, the implementation and the testing of that algorithm with data sets coming from regular customers of Clear Destination, (ii) the design of a machine learning model and algorithm for the computation of the size of the time windows based on the type of goods and the characteristics of the pickup/delivery addresses, using the current available data, its implementation and testing , and (iii) a set of recommendations for the collection or the acquisition of more data related to the characteristics of the pickup/delivery addresses.
Benefit for the partner organization will be a more flexible and robust software for the planning of last mile pickup and delivery, with shorter and more accurate time windows. This will in turn increase the satisfaction of the destined clients of the pickup and delivery operations.

Jean Lucas de Sozua Toniolli
Faculty Supervisor: 
Brigitte Jaumard
Partner University: