Performance of mechatronics systems can deteriorate significantly under the presence of delays in measurements. On the other hand, intentional introduction of a delay in a feedback control law may be beneficial in achieving good performance without relying on accurately tuned estimates for derivatives of measured signals. In this context, a proper design of retarded control laws represents an important challenge, and this paper gives a step forward to obtain better control laws for underactuated mechanical systems under the presence of time delays in measurements. Particularly, a Proportional-Retarded (PR) control law is designed for stabilization of an underactuated rotational inverted pendulum, known as the Furuta pendulum. Experiments over a laboratory platform as well as a comparison with a Linear Quadratic Regulator (LQR) are performed to show the advantages of the proposed scheme.

Due to some limitations of mechanical solution for automation of front end loaders, specifically self-leveling task, we investigate here the sensor-based solution. We present a model of front end loaders and the comparison of different controllers to realize this task. The best performance is observed from the controller comprising linear terms, compensation of input nonlinearity known as dead zone, and compensation of disturbances constructed from its estimation.