AutoPack 2.0 – Automatic packaging and assembly of electrical cables based on optimization and machine learning

AutoPack 2.0 wants to increase the degree of digitization and automation in the product- and production development process and to ensure that production and assembly aspects are taken into consideration early in the development process. The application of the project is packing and assembly of high-voltage cables and cable harnesses, which is a time-consuming and resource-intensive activity throughout the automotive industry. With the electrification of the vehicle fleet, the experience built up around the installation of hoses and cables for internal combustion engines needs to be renewed based on the new conditions that electrification entails. In order to meet these new requirements and to take production and assembly aspects into account in the early stages of development, both new methods and tools need to be developed. The proposed project is a direct continuation of the AutoPack project, financed via FFI-Machine Learning.

To define hoses, pipes and cabling installations at Volvo, tens of thousands of engineering hours are spent at a cost of tens of million SEK per year. With the help of further developing the Autopack toolbox, it is estimated that AutoPack 2.0 will deliver a 50% reduction in the engineering hours spent on packaging, production planning and assembly, and the corresponding cost savings.