Machine learning for optimized and reliable storage of grain
| Reference number | |
| Coordinator | SKANDIA ELEVATOR AB |
| Funding from Vinnova | SEK 500 000 |
| Project duration | October 2019 - November 2020 |
| Status | Completed |
| Venture | AI - Competence, ability and application |
| Call | Start your AI journey! |
Purpose and goal
The purpose has been to build knowledge within Skandia Elevator (SE) about AI / ML, evaluate the usefulness of the technology to create a calculation algorithm and demonstrate the results internally within the company. Within SE product and production-related departments, there are today several ideas where AI / ML could be suitable to use for problem-solving. The project´s goals are considered achieved.
Expected results and effects
ML was applied to a product that in parallel with the project developed a calculation algorithm in a classic way. Those concerned within the company could relate to the chosen product and put methodology and technology in a known context and thus realize its possibilities and limitations. The ML-based algorithm is a strong candidate to use for the upcoming product´s calculations. Final verification is still ongoing (Nov. 2020).
Planned approach and implementation
On data collected during the autumn of 2019, before the start of the project, a handful of calculation algorithms (candidates) were developed based on different ML methods by the project participants. The candidates were evaluated Sept / Oct 2020 with data collected during the harvest in 2020. The best algorithms gave at least as good results as the "classic" algorithm while it was simpler, with fewer variables. AI / ML training was conducted for Skandia Elevator for one day in October 2020, during which several new innovation ideas emerged.