Automated monitoring of industrial cleaning processes - DigiClean II

Important results from the project

The project has developed and validated a sensor and AI platform that replaces manual lab analysis with real-time monitoring of washing liquids. Machine learning models reach prediction errors of around 2% and a model is deployed with automatic dosing at an industrial partner. Tests were conducted at Volvo Group GTO Skövde. A two-stage architecture enables scaling across facilities. A service-based business model was developed and the company DigiClean Solutions AB was established as a spin-off.

Expected long term effects

In the long term, the project is expected to reduce the consumption of chemicals, water and energy in industrial washing and extend the life of baths, which strengthens the sustainability of Swedish industry. The results enable a shift from volume-based sales to a service-based business model. The technology can be scaled to water treatment, food, pulp/paper and electronics. DigiClean Solutions AB, an expanded network and EIT visibility provide a basis for continued innovation and investments.

Approach and implementation

The project was implemented by a consortium of RISE, CIT, IKP AB, Clemondo AB, SKF, Proton AB and DigiClean AB, in three work packages: platform development, data collection and modeling, and business model. An agile, iterative working method was applied with close collaboration between technology and business development. Field tests were carried out at Volvo GTO provided valuable industrial data. The schedule was followed and the cooperation in the consortium worked well through the project