Population growth, climate change, resource depletion and soil degradation are placing huge challenges on agriculture. One way to alleviate these problems could be to increase the degree to which robots are used in agricultural settings. Rigid robotic sensing and actuation for repeatable tasks in industrial settings is well-developed, but applying robotics in agriculture remains challenging, because fruits and vegetables damage easily, and agricultural environments are unstructured. Soft robots could be a solution, since they are naturally suited to replicate the soft touch that humans normally use to for instance harvest a raspberry. Moreover, their soft structure creates an inherent adaptability to for instance fruits with different sizes. However, soft force sensors still perform much worse than rigid force sensors do.
In this PhD project, you will explore two related approaches to enhance soft sensor performance: co-design of sensors&actuators and active sensing. For robotic design, the measurement principle of the sensor is taken as the starting point, and then the actuator is co-designed to optimize the functionality of the sensor. For control, active sensing strategies, in which the sensor is intentionally moved around to gather data, will be used to continuously update the robot's movements to collect the most useful data. The final goal is to run field experiments with real autonomous robots.
You will become a member of the Robotics research line of the
Control Systems Technology section within the Mechanical Engineering department of the TU/e. This project will also be part of the HTSF Green Sensor Program, which is a collaboration of the 4 technical universities of the Netherlands. Furthermore, you will be a part of the Reshape lab, which has state-of-the art facilities for manufacturing and testing of soft robots.
Within the project, there is a lot of room for your personal input. Some of the techniques that will probably be part of the project are:
- Rapid prototyping using CAD design and 3D printing
- Silicone casting of soft robots
- (Simple) electronics development
- Controller design for real-time control, for instance with Python
- Running validation experiments with commercially-available force sensors
- Data analysis with for instance Matlab/Python/R
- Field experiments with real robots