EnBlightMe! - an automated support system for potato late blight detection
|Coordinator||Sveriges Lantbruksuniversitet - Inst för växtskyddsbiologi|
|Funding from Vinnova||SEK 1 574 360|
|Project duration||December 2016 - October 2019|
Purpose and goal
In Sweden one fourth of all fungicides is used against potato late blight. EnBlightMe! aimed to automatically detect late blight in fields with the help of multispectral analysis, computer vision and drones via a prototype app to lower pesticide use. The app considered other useful input for decision-making such as weather and economy data. A demo-app was successfully developed and demonstrated in collaboration with IBM. Ideas were tested for precision agriculture, organic farming and breeding.
Expected results and effects
The work was communicated at an end conference and in numerous news features and articles. It was also captured as part of a peer-reviewed review article. The work led to the successful grant application NordPlant (www.nordplant.ord) and has strengthened the collaboration with the Nordic PPP project for plant phenotyping for breeding. We envisage further development around satellite analysis and computer vision to combat plant disease.
Planned approach and implementation
EnBlightMe! was initiated in 2017 with a DesignThink workshop led by IBM with invited researchers, potato growers and software developers. It had a clear focus on the users of the solution and their specific needs and formed the base for the project. Most of the work was then performed in the work package groups. Both MSc and project students were involved. In November 2018, the demo-app and results were presented with many of the participants during the initial workshop present.