Flavio Roberti

Direct visual tracking control of remote cellular robots

Abstract This paper presents the design of a stable non-linear control system for the remote visual tracking of cellular robots. The robots are controlled through visual feedback based on the processing of the image captured by a fixed video camera observing the workspace. The control algorithm is based only on measurements on the image plane of the visual camera–direct visual control–thus avoiding the problems related to camera calibration. In addition, the camera plane may have any (unknown) orientation with respect to the robot workspace. The controller uses an on-line estimation of the image Jacobians. Considering the Jacobians’ estimation errors, the control system is capable of tracking a reference point moving on the image plane–defining the reference trajectory–with an ultimately bounded error. An obstacle avoidance strategy is also developed in the same context, based on the visual impedance concept. Experimental results show the performance of the overall control system.

Stable contour-following control of wheeled mobile robots

This paper presents a continuous wall-following controller for wheeled mobile robots based on odometry and distance information. The reference for this controller is the desired distance from the robot to the wall and allows the robot to follow straight wall contour as well as smoothly varying wall contours by including the curvature of the wall into the controller. The asymptotic stability of the control system is proved using a Lyapunov analysis. The controller is designed so as to avoid saturation of the angular velocity command to the robot. A novel switching scheme is also proposed that allows the robot to follow discontinuous contours allowing the robotic system to deal with typical problems of continuous wall-following controllers such as open corners and possible collisions. This strategy overcomes these instances by switching between dedicated behavior-based controllers. The stability of the switching control system is discussed by considering Lyapunov concepts. The proposed control systems are verified experimentally in laboratory and office environments to show the feasibility and good performance of the control algorithms.

Nonlinear Control Techniques and Omnidirectional Vision for Team Formation on Cooperative Robotics

In this work a robot cooperation strategy based on omnidirectional vision is presented. Such strategy will be applied to a mobile robot team formed by small and simple robots and a bigger leader robot with more computational power. The leader must control team formation. It has an omnidirectional camera and sees the other robots. Color segmentation and Kalman filtering is used to obtain the pose of the followers. This information is then used by a nonlinear stable controller to manage team formation. Simulations and some preliminary experiments were run. The current results are interesting and encourage towards the next steps.

Vision-Based Control of the RoboTenis System

In this paper a visual servoing architecture based on a parallel robot for the tracking of faster moving objects with unknown trajectories is proposed. The control strategy is based on the prediction of the future position and velocity of the moving object. The synthesis of the predictive control law is based on the compensation of the delay introduced by the vision system. Demonstrating by experiments, the high-speed parallel robot system has good performance in the implementation of visual control strategies with high temporary requirements.

Robust control with redundancy resolution and dynamic compensation for mobile manipulators

This paper presents the tracking control problem of a mobile manipulator system to maintain maximum manipulability and including the obstacle avoidance. The design of the controller is based on two cascaded subsystems: a minimum norm kinematic controller with command saturation, and a controller that compensates for the dynamics of the mobile manipulator system. Robot commands are defined in terms of reference velocities. Stability and robustness to parametric uncertainties are proved by using Lyapunovs method. Experimental results show a good performance of the proposed controller as proved by the theoretical design.

Switched Control to Robot-Human Bilateral Interaction for Guiding People

This paper presents a switched control strategy to interpret and design a human-robot bilateral interaction when a human follows a non-holonomic mobile robot at a desired distance while the robot is already following a known path. Furthermore, it proposes and experimentally validates a model that mathematically describes the human behavior when performing the specific task of tracking a mobile robot. This model is useful for the purposes of the control system design and its associated stability analysis. A switched system is proposed to model the complete human-robot behavior. The switching strategy is based on the human-robot relative position and on the human intention to follow the robot. Control errors are defined in terms of human to robot and robot to path instantaneous distances. Stability analyses for the individual controllers, as well as for the complete switching system, are provided by considering Lyapunov theory. Real human-robot interaction experiments show the good performance of the proposed control strategy.

Robust Control with Dynamic Compensation for Human-Wheelchair System

This work presents the kinematic and dynamic modeling of a human-wheelchair system, and dynamic control to solve the path following problem. First it is proposed a dynamic modeling of the human-wheelchair system where it is considered that its mass center is not located at the center the wheels’ axle of the wheelchair. Then, the design of the control algorithm is presented. This controller design is based on two cascaded subsystems: a kinematic controller with command saturation, and a dynamic controller that compensates the dynamics of the robot. Stability and robustness are proved by using Lyapunov’s method. Experimental results show a good performance of the proposed controller as proved by the theoretical design.

Adaptive cooperative control of multi-mobile manipulators

This paper presents a multi-layer scheme for the adaptive coordinated cooperative control of n mobile manipulator. Each layer works as an independent module, dealing with a specific part of the problem of cooperation, coordination and adaptation, thus giving more flexibility to the scheme. Also, the redundancy of the n mobile manipulators is used for the avoidance of obstacles by the mobile platforms-without deforming the virtual structure and maintaining its desired trajectory- and the singular configuration prevention through the systems manipulability control. Stability and robustness are proved by using Lyapunovs theory. Simulation results show a good performance of the proposed control scheme as proved by the theoretical design.

Control Servo-Visual de un Robot Manipulador Planar Basado en Pasividad

In this work a visual servo control based on the passivity properties of the visual system is designed. It proposes a regulator with variable control gains, to avoid the saturations of the actuators while introducing the ability to correct errors of small magnitude. Also the design is done taking into account the L2 performance, to give the capacity to track moving objects, with a small error control. Experimental results are showed in an industrial robot manipulator planar type to verify the compliance with the objectives of the proposed controller.

One camera in hand for kinematic calibration of a parallel robot

The main purpose of robot calibration is the correction of the possible errors in the robot parameters. This paper presents a method for a kinematic calibration of a parallel robot that is equipped with one camera in hand. In order to preserve the mechanical configuration of the robot, the camera is utilized to acquire incremental positions of the end effector from a spherical object that is fixed in the word reference frame. Incremental positions of the end effector are related to incremental positions of encoders of the motors of the robot. A kinematic model of the robot is modified in order to take into account possible errors of kinematic parameters. The solution of the model utilizes incremental positions of the resolvers and end effector, the new parameters minimizes errors in the kinematic equations. Spherical properties and intrinsic camera parameters are utilized to model sphere projection in order to improve spatial measurements. The robot system is designed to carry out tracking tasks and the calibration of the system is finally validated by means of integrating the errors of the visual controller.