AI
Pilot Project: Students Protecting Apples Using IoT

Detecting pests is crucial for farmers but checking the pest traps is time-consuming. This is why a group of students created a smart solution that enables the farmers to focus on what matters – their apples.

Published : 04.06.2020

Pests are a common nuisance for farmers. Detecting pests is crucial and thus checking the pest traps is a daily chore that is time-consuming. This is why a group of the IoT Rapid-Proto Labs students created a smart solution that enables the farmers to focus on what matters – their apples. The pest detection system is already up and running in chosen apple orchids, red fruit fields and wineries in Italy. The results are promising.

Camera: the foundation for smart pest detection

In the early prototype of the pest detection system a camera was placed inside a pest trap. The camera took pictures daily and sent them to the farmer to analyze which insects were there and if there is need for actions.

The IoT Rapid-Proto Labs student group added intelligence to the camera. They created a model where the camera understands whether the pest is dangerous or not. When the camera classifies an insect as harmful, the farmer gets a notification. The statistics tell the farmer when the insect population needs actions.

To improve the accuracy of the pest detection system, sensors were set to measure the vibration insects make when flying. Every insect species has their very own frequency of flapping wings. Combining the sensor data with data provided by the camera, the classification becomes even more accurate.

The importance of energy autonomy

If the pest detecting cameras would not be energy sufficient, the farmers should be changing batteries every once in a while, which would then ruin the entire idea of not having to check the pest traps. One job (checking the pest traps) would be just replaced with a new one (changing the batteries).
The IoT Labs students solved the problem by using a low-power wide-area network protocol, also known as LoRaWAN. As it is license free and wireless with a range of 15 kilometers, it is more suitable as there is no need to maintain WiFi or Bluetooth connections.

The original pest detection solution will be developed by a new student group continuing the work where the previous group finished. In addition to pest detection, the same technology is applied in a surveillance solution by another student team. The solution may even offer the possibility for entrepreneurship as some of the students have plans on creating a startup based on it. The possibilities with IoT are endless.

Authors: professor Davide Brunelli, University of Trento and project coordinator Nella Kelkka, Haaga-Helia University of Applied Sciences

The writers are experts in the IoT Rapid-Proto Labs project that creates new IoT talent in Europe. The project partners are University of Trento, Leiden University, TU Delft, Haaga-Helia University of Applied Sciences, Fondazione Bruno Kessler, 247Grad and Houston Inc. The project is co-funded by Erasmus+.