Data-Driven Model Predictive Control for Food-Cutting
This work addresses the problem of robotic food-cutting for automation applications, representing an incremental improvement by extending earlier torque-controlled methods to velocity-controlled robots with sensor integration.
The paper tackles the challenge of controlling contact-rich food-cutting tasks by developing a data-driven model predictive control approach using a recurrent neural network to model dynamics, achieving efficient cutting rates on both trained and unknown objects with a single dynamic model.
Modelling of contact-rich tasks is challenging and cannot be entirely solved using classical control approaches due to the difficulty of constructing an analytic description of the contact dynamics. Additionally, in a manipulation task like food-cutting, purely learning-based methods such as Reinforcement Learning, require either a vast amount of data that is expensive to collect on a real robot, or a highly realistic simulation environment, which is currently not available. This paper presents a data-driven control approach that employs a recurrent neural network to model the dynamics for a Model Predictive Controller. We build upon earlier work limited to torque-controlled robots and redefine it for velocity controlled ones. We incorporate force/torque sensor measurements, reformulate and further extend the control problem formulation. We evaluate the performance on objects used for training, as well as on unknown objects, by means of the cutting rates achieved and demonstrate that the method can efficiently treat different cases with only one dynamic model. Finally we investigate the behavior of the system during force-critical instances of cutting and illustrate its adaptive behavior in difficult cases.