ECSEL 2017 RIA aFarCloud Imagimob
| Reference number | |
| Coordinator | Imagimob AB |
| Funding from Vinnova | SEK 1 896 882 |
| Project duration | September 2018 - December 2021 |
| Status | Completed |
Purpose and goal
Within agriculture, fewer and fewer people need to do more work, and the farmer needs decision-making support to be able to control seed, irrigation, fertilization, spraying and animal holding. Within aFarCloud, a cloud platform has been developed for the collection, processing and visualization of data. Imagimob has integrated a solution for activity measurement for animals and agricultural equipment, where motion data is processed sensor-near, and sent to the cloud for visualization. We hold an automated process for data collection, and training of neural networks
Expected results and effects
We have developed an equivalent for livestock of a sports tracking bracelet, which works over long distances. Tests of the services we have developed show that the animals´ activity and movement patterns can be predicted with high probability, and give an indication of an animal´s well-being over time. This can provide savings for the animal farmer, who by being able to continuously monitor the animals, can see long-term effects of following a meat or dairy cow for longer periods to be able to early detect hidden underlying ailments and prevent infection spreading to the rest of the herd.
Planned approach and implementation
The work has undergone several phases; collection, training of AI models, deployment, testing and lastly feedback and hence new iterations. Data collection from tractors and other fixed assets has been straight-forward, while data from animals is more demanding, often in combinations of monitored collection (video), supplemented with unattended methods. Exchange of knowledge and harmonization between partners working groups´ methods also allowed other actors to be able to use our tools for analysis of their own data, whereby economies of scale have been achieved in the project.