Deep Process Learning
| Reference number | |
| Coordinator | RISE Research Institutes of Sweden AB - SICS Swedish ICT Västerås AB |
| Funding from Vinnova | SEK 2 908 000 |
| Project duration | May 2017 - April 2019 |
| Status | Completed |
Important results from the project
This project aims to show how deep learning can be used to introduce a big leap for automation in process industry. The proposed approach will take advantage of the data already gathered in the process control system and use that to suggest the needed action to improve desired KPI:s, contributing to optimize quality of the final product.
Expected long term effects
The results of the project can in the first step be used as a decision support for operators in the pulp and paper industry. The intention and goal to completely automate the control of the process has not been achieved within the project. An important result is increased knowledge about present possibilities and limitations of deep learning and how the technique can be applied in process industry.
Approach and implementation
The project has been conducted in close cooperation between the parties, with a delegated responsibility for the different work packages: Data collection, Development of deep learning algorithm for process industry, User experience and visualization, and Dissemination. An important part of the project was to collect large amounts of data, especially about deviations, and to secure quality of data and to establish ways to transfer data in a robust and secure way.