
The project aims to gradually develop and test new software solutions in collaboration with those working in the production environment. By involving operators and experts actively in the development process, both their understanding and trust in the predictions made by the systems increase, something that in turn can lead to better decision-making and more accurate maintenance work.
Efficient maintenance within production systems is a critical area for industry in order to avoid costly stops in production as well as achieving maximal utilisation of parts/components of machines.
One way to approach such maintenance work is to gather data from sensors, placed on the machines, and build models based on the data which are able to predict when or how long additional time a component can be utilised.
However, these models, often based on different types of artificial intelligence, can be hard to understand, and their predictions can also be associated with rather much of uncertainty, which can make it complicated for operators to build trust towards the predictions.
The goal of this project is to iteratively develop software prototypes which can be tested and evaluated in close collaboration with operators or experts in industrial production systems.
Such development procedures, where operators are an active part of the development, are expected to increase their trust and understanding of the predictions and thereby constitute a better foundation for decision-making regarding maintenance operations.
Project management
Chalmers University of Technology is leading the project. The Project Manager for the University of Skövde’s part of the project is Alexander Karlsson.