The stawberry agricultural industry is growing worldwide. However, the crop has several challenges: a short window of ripeness, sensitivity to weather and climate, and vulnerability to pests and rot. These issues may be mitigated by carefully picking each berry at its optimal state of ripeness, and quickly removing diseased berries. Robotic strawberry pickers are a promising technology for achieving these goals; however, current pickers lack the advanced perception needed to evaluate the quality of each berry.
The MålBær project will address these shortcomings by developing sensor hardware for assessing and tracking the ripeness and health of each strawberry. All measurements must be non-contact to avoid spreading of disease, and sensors must be miniaturized for integration with the robotic picker. Computer analytics will be developed for merging and converting raw sensor data into a set of quality measurements, which the robot can use to decide if a berry is ready for picking or otherwise generate a forecast for when it will reach maturity.
SINTEF's development work will involve the design and realization of experimental hardware, development of software algorithms, and field tests in real agricultural settings. In the second year of the project, we will merge the sensors with SAGA Robotics' robot platform Thorvald and test the system through the full lifecycle of the strawberries.