This customer is one of the largest rental equipment providers in the Nordics and supplies thousands of construction sites. They needed a system capable of predicting when rented-out equipment was expected to be returned.
The model takes as input detailed characteristics of the equipment, the customer, geography, and season and estimates when equipment will be returned. This acts as a cornerstone for planning its vast logistics chain and minimizing unnecessary transports between its local shops and the main warehouses.
Initial testing indicated savings in the order of millions of Euros per year and a positive climate impact due to a reduction in transports.
They are one of the largest of their kind in Europe and provide hundreds of thousands of pieces of equipment and supply tens of thousands of construction sites. Their catalog contains everything from small hand-held tools to scaffolding and large diesel generators.
How Modulai did it
In close collaboration with the company’s data scientists, the Modulai team developed a pipeline for preprocessing data from their large database of customers, construction projects, and past rentals. The team developed and assessed many regression and classification approaches. The system was delivered as a Python service to be run on their internal infrastructure and integrated with the ERP system.