Robots are traditionally designed for motion control tasks. This has resulted in massive and rigid structures with high positioning accuracy. These robots are designed for tasks that require accurate positioning of the robot’s end-effector; where the forces experienced by the end-effector are treated as disturbances and compensated for by the tight position control algorithms. The rigid structures do not provide compliance (softness) because compliance may result in motion inaccuracies. Such robots have been very useful for tasks where the robot is not in contact with its environment especially when carrying payloads in pick and place operations. Furthermore, these robots operate in well-structured and engineered environments where the geometry of the environment is exactly known. Any inaccuracies of workpiece locations cannot be tolerated.
Therefore, industrial robots are usually bulky and have high inertia, and this limits their responsiveness in terms of mechanical/control bandwidth. An add-on (“mini”) manipulator is one way to help improve system performance without having to re-design the existing manipulator. This leads to the “macro/mini” manipulator concept, where the add-on “mini” manipulator is attached to the end of the “macro” manipulator (i.e. the industrial robot). This mini manipulator will be designed in-house such that it is optimized both mechanically (e.g. low inertia and fast response) and via advanced control schemes. Then, the macro and the mini manipulators are combined to offer the best of both worlds: the large workspace of the macro, and the fast response as well as good precision of the mini.
This work package WP4 aims to equip industrial robots with compliant motion capabilities to do tasks that require continuous contact with the environment; and the capabilities will be developed further to address 3D surface polishing. These are realized by the following three objectives:
Real-time strategies and algorithms for the planning and control of the motion of a robotic manipulator to achieve the desired force, motion and compliant motion profiles, where the end-effector is in continuous contact with the environment (workpiece) whose configuration is not exactly known. These low-level control strategies are more useful if they are tightly integrated to the higher level planning strategies that provide the desired force and compliant motion profiles. Depending on the realtime performance of the algorithms and task execution quality, the desired profiles can be adjusted in real-time to improve quality and productivity.
Design a framework for the control of a macro/mini manipulator to improve the performance of the manipulator especially in force control and finishing applications. The research questions to be answered in this project include cross coupling effects and vibrations, high level planning for cooperation between the macro and mini manipulators, optimal control strategy based on macro/mini configuration and controllable DOF, as well as control issues due to different sampling rate and accessibility level of the two counterparts.
Develop strategies and algorithms for robotic finishing (investigated in WP7) of metal surfaces that require uneven removal rates. Polish finishing is a critical stage in many product manufacturing processes. It is to remove material by polishing the uneven surface with an unknown/uncertain profile. Current practice is often by manual operation because surface profile uncertainties prevent preprogramming a robot to carry out such a task. Theory of position and force control for robot constrained motion has been extensively studied for three decades. However, successful industrial applications of robot with force control are still far and few for 3D surface polishing.