The growth of robotics is mainly driven by the expansion into new application areas. Existing robotic applications typically require high forces, high precision and are highly repetitive. Therefore, conventional robots are usually made of high-stiffness material such as steel or aluminum, equipped with strong actuators, and designed for a very specific task. However, this also results in a large risk potential. This limits their application to new fields such as healthcare, humane-machine interaction or gripping of fragile and differently shaped objects.
One approach to closing this gap is the use of soft robots. Soft robots are usually made of soft materials such as silicone or foam. This allows large elastic deformations, limits the occurring forces and minimizes their risk potential. At the same time, very different objects can be gripped with one and the same soft gripper, since soft robots can easily adapt to different shapes.
The goal of our research is to improve the supervision and control of soft grippers by integrating gripping force and shape sensors. Additionally, we aim to improve the agility and accuracy of soft robots by developing modern dynamic control methods for soft robots.
Due to the soft structure, conventional components and design methodologies are not applicable. New actuators, sensors and control concepts are currently developed. Conventional actuators cannot be integrated into soft material robots, since they are rigid components. They would counteract the soft structure. Therefore, alternative concepts, such as cable-driven actuators, shape memory alloys (SMA) or hydraulic actuators are currently developed. For reliable control and positioning of the robot in 3D space, not only actuation is important but accurate knowledge about its current position and orientation is even more essential. However, conventional sensors cannot be used for estimation of the robots curvature, since they would destroy the robots softness.
New concepts must also be developed for the modeling and control of soft robots. In contrast to rigid robots and flexible robots, large elastic deformations occur in soft robots. Therefore, established modeling methods in robotics, such as modeling with flexible multibody systems, are unsuitable for soft robots.
Sensor feedback is essential for precise trajectory tracking and advanced gripping applications. In particular, force sensors are needed to control the gripping force and shape sensors are needed to measure the robot configuration. The main challenges in developing sensors for soft robots are the large strains that occur in soft robots and that the sensors must be soft.
For supervision and control of complex gripping processes the integration of different sensors into the gripper is essential.
This topic offers a combination of mechanics and control engineering and provides much space for student work. Please contact us, if you are interested to contribute to this topic within the scope of a project work or a Bachelor-/Masterthesis.