ALISTAIR - Artificial Intelligence for Sustainable Production

Important results from the project

The goal was to develop a method for addressing GHG emissions in production plants, and to apply it in Diamyd Medical´s facility. The project has catalyzed both an AI and data perspective on the production facility as well as the sustainability perspective during the build-up of the facility including mapping out the production process, recyclability, energy consumption, warehousing, waste management and process flow. Significant optimizations have been made in ventilation systems of the facility to minimize the energy use and provide detailed data of the energy use of the facility.

Expected long term effects

The most important goal that has been achieved is that Diamyd Medical´s facility is much more optimized in regard of the sustainability aspect than it was before the project started, which includes employee awareness. Several data streams have been identified and data has been used for AI-based optimization by MainlyAI. KTH has been instrumental in circular economy use cases such as waste management in production industry.

Approach and implementation

MainlyAI has developed and trained a Neural Network for predicting temperature and humidity in the facility. Diamyd Medical have installed temperature and humidity sensors in the facility. Data is collected and connected to a cloud server using a dedicated local wireless network. The data has been used by MainlyAI to predict temperature and humidity in any of the locations were the sensors are located (where data has been collected). The predictions can be used to better control the climate of the facility, save energy, and to avoid too high/low temperature/humidity in the facility.