Paolo Boscariol

Trajectory Planning in Robotics

Trajectory planning is a fundamental issue for robotic applications and automation in general. The ability to generate trajectories with given features is a key point to ensure significant results in terms of quality and ease of performing the required motion, especially at the high operating speeds necessary in many applications. The general problem of trajectory planning in Robotics is addressed in the paper, with an overview of the most significant methods, that have been proposed in the robotic literature to generate collision-free paths. The problem of finding an optimal trajectory for a given path is then discussed and some significant solutions are described.

Model Predictive Control of a Flexible Links Mechanism

Vibration suppression in flexible link manipulator is a recurring problem in most robotic applications. Solving this problem would allow to increase many times both the operative speed and the accuracy of manipulators. In this paper an innovative controller for flexible-links mechanism based on MPC (Model Predictive Control) with constraints is proposed. So far this kind of controller has been employed almost exclusively for controlling slow processes, like chemical plants, but the authors’ aim is to show that this approach can be successfully adapted to plants whose dynamical behavior is both nonlinear and fast changing. The effectiveness of this control system will be compared to the performance obtained with a classical industrial control. The reference mechanism chosen to evaluate the effectiveness of this control strategy is a four-link closed loop planar mechanism laying on the horizontal plane driven by a torque-controlled electric actuator.

Experimental Validation of Minimum Time-jerk Algorithms for Industrial Robots

In this paper, we present a minimum-time/jerk algorithm for trajectory planning and its experimental validation. The algorithm search for a trade-off between the need for a short execution time and the requirement of a sufficiently smooth trajectory, which is the well known necessary condition to limit the vibration during fast movements. The trade-off is achieved by adjusting the weight of two suitable functions, able to consider both the execution time and the squared-jerk integral along the whole trajectory. The main feature of this algorithm is its ability to smooth the trajectory’s profile by adjusting the intervals between two consecutive via-points so that the overall time is minimally delayed. The practical importance of this technique lies in the fact that it can be implemented in any industrial manipulator without a hardware upgrade. The algorithm does not need for a dynamic model of the robot: only the mechanical constraints on the position, velocity and acceleration ranges have to be set a priori. The experimental proof is provided in this paper by comparing the results of the proposed algorithm with those obtained by adopting some classical algorithms.

Abigaille-III: A Versatile, Bioinspired Hexapod for Scaling Smooth Vertical Surfaces

This paper presents a novel, legged robot, Abigaille-III, which is a hexapod actuated by 24 miniature gear motors. This robot uses dual-layer dry adhesives to climb smooth, vertical surfaces. Because dry adhesives are passive and stick to various surfaces, they have advantages over mechanisms such as suction, claws and magnets. The mechanical design and posture of Abigaille-III were optimized to reduce pitchback forces during vertical climbing. The robot’s electronics were designed around a Field Programmable Gate Array, producing a versatile computing architecture. The robot was reconfigured for vertical climbing with both 5 and 6 legs, and with 3 or 4 motors per leg, without changes to the electronic hardware. Abigaille-III demonstrated dexterity through vertical climbing on uneven surfaces, and by transferring between horizontal and vertical surfaces. In endurance tests, Abigaille-III completed nearly 4 hours of continuous climbing and over 7 hours of loitering, showing that dry adhesive climbing systems can be used for extended missions.

Optimal Gait for Bioinspired Climbing Robots Using Dry Adhesion: A Quasi-Static Investigation

Legged robots relying on dry adhesives for vertical climbing are required to preload their feet against the wall to increase contact surface area and consequently maximize adhesion force. Preloading a foot causes a redistribution of forces in the entire robot, including contact forces between the other feet and the wall. An inappropriate redistribution of these forces can cause irreparable detachment of the robot from the vertical surface. This paper investigates an optimal preloading and detaching strategy that minimizes energy consumption, while retaining safety, during locomotion on vertical surfaces. The gait of a six-legged robot is planned using a quasi-static model that takes into account both the structure of the robot and the characteristics of the adhesive material. The latter was modelled from experimental data collected for this paper. A constrained optimization routine is used, and its output is a sequence of optimal posture and motor torque set-points.

Path Planning and Trajectory Planning Algorithms: A General Overview

Path planning and trajectory planning are crucial issues in the field of Robotics and, more generally, in the field of Automation. Indeed, the trend for robots and automatic machines is to operate at increasingly high speed, in order to achieve shorter production times. The high operating speed may hinder the accuracy and repeatability of the robot motion, since extreme performances are required from the actuators and the control system. Therefore, particular care should be put in generating a trajectory that could be executed at high speed, but at the same time harmless for the robot, in terms of avoiding excessive accelerations of the actuators and vibrations of the mechanical structure. Such a trajectory is defined as smooth. For such reasons, path planning and trajectory planning algorithms assume an increasing significance in robotics. Path planning algorithms generate a geometric path, from an initial to a final point, passing through pre-defined via-points, either in the joint space or in the operating space of the robot, while trajectory planning algorithms take a given geometric path and endow it with the time information. Trajectory planning algorithms are crucial in Robotics, because defining the times of passage at the via-points influences not only the kinematic properties of the motion, but also the dynamic ones. Namely, the inertial forces (and torques), to which the robot is subjected, depend on the accelerations along the trajectory, while the vibrations of its mechanical structure are basically determined by the values of the jerk (i.e. the derivative of the acceleration). Path planning algorithms are usually divided according to the methodologies used to generate the geometric path, namely: roadmap techniques cell decomposition algorithms artificial potential methods.

Active Position and Vibration Control of a Flexible Links Mechanism Using Model-Based Predictive Control

In order to develop an efficient and,fast position control for robotic manipulators, vibration phenomena have to be taken into account. Vibrations are mainly caused by the flexibility of manipulator linkages, especially when dealing with high-speed and lightweight robots. In this paper, a constrained model-based predictive control is employed for controlling both position and vibrations in a mechanism with high link flexibility. This kind of controller has so far been used mainly to control slow processes, but here simulation results that show its effectiveness in dealing with high-speed and nonlinear processes are presented. The mechanism chosen to evaluate the performances is a four-link closed chain mechanism laying on the horizontal plane and driven by a single torque-controlled electric motor.

Design of a controller for trajectory tracking for compliant mechanisms with effective vibration suppression

In this paper, a numerical investigation of the Model Predictive Control strategy applied to flexible-link mechanisms is presented. The mechanisms used for all the tests are a planar five-link mechanisms. The tests are aimed at showing how the proposed control system can be used for the trajectory tracking and the vibration suppression. An analysis of the effects of the choice of tuning parameters is presented as well. The design of the predictive controller is based on a linearized version of an accurate nonlinear dynamic model. The effectiveness of the proposed approach is confirmed by extensive numerical results.

Fabrication and performance analysis of a DEA cuff designed for dry-suit applications

A method for manufacturing a cylindrical dielectric elastomer actuator (DEA) is presented. The cylindrical DEA can be used in fabricating the cuff area of dry-suits where the garment is very tight and wearing the suit is difficult. When electrically actuated, the DEA expands radially and the suit can be worn more comfortably. In order to study the performance of the DEA, a customized testing setup was designed, and silicone-made cuff samples with different material stiffnesses were tested. Analytical and FEM modeling were considered to evaluate the experimental output. The results revealed that although the stiffness of the DEA material has a direct relationship with the radial constrictive pressure caused by mechanically stretching the DEA, it has a minor effect on the actuation pressure. It was also found that stacking multiple layers of the DEA to fabricate a laminated structure enabled the attainment of a desired variation of pressure required for the implementation of an electrically tunable cuff.

Vibration reduction in a flexible link mechanism through the synthesis of an MPC controller

The modeling and the control of flexible link robots have received a great deal of attention in the last decades due to the wide prospective industrial and space applications of ultra-light and high-speed mechanisms. This paper introduces a general and practical procedure for the design of effective control schemes for the position and vibration control of flexible links mechanisms. In particular, an innovative controller based on MPC (Model-based Predictive Control) is proposed. So far the MPC controllers have been employed almost exclusively in slow industrial processes. Nevertheless, this work shows that the MPC approach can be successfully adapted to plants whose dynamics are both nonlinear and fast changing as well. The performances of this approach will be evaluated for a single link mechanism and compared to those obtained with a standard PID position controller.