Abstract:Flexible manipulators based on soft robotic technologies demonstrate compliance and dexterous maneuverability with virtually infinite degrees-of-freedom. Such systems have great potential in assistive and surgical fields where safe human-robot interaction is a prime concern. However, in order to enable practical application in these environments, intelligent control frameworks are required that can automate low-level sensorimotor skills to reach targets with high precision. We designed a novel motor learning a… Show more
“…There are also hybrid versions such a cell-based Voronoi roadmap generation algorithm that is searched with A* [10]. Attempts have been made with reinforcement learning [29,30]. All of these algorithms are in most of the cases based on high-dimensional configuration spaces, corresponding to the robot joint positions.…”
This paper argues that an efficient artificial intelligence control algorithm needs the built-in symmetries of an industrial robot manipulator to be further characterized and exploited. The product of this enhancement is a four-dimensional (4D) discrete cylindrical grid space that can directly replace complex robot models. A* is chosen for its wide use among such algorithms to study the advantages and disadvantages of steering the robot manipulator within the 4D cylindrical discrete grid. The study shows that this approach makes it possible to control a robot without any specific knowledge of the robot kinematic and dynamic models at planning and execution time. In fact, the robot joint positions for each grid cell are pre-calculated and stored as knowledge, then quickly retrieved by the pathfinding algorithm when needed. The 4D cylindrical discrete space has both the advantages of the configuration space and the three-dimensional Cartesian workspace of the robot. Since path optimization is the core of any search algorithms, including A*, the 4D cylindrical grid provides for a search space that can embed further knowledge in form of cell properties, including the presence of obstacles and volumetric occupancy of the entire industrial robot body for obstacle avoidance applications. The main trade-off is between a limited capacity for pre-computed grid knowledge and the path search speed. This innovative approach encourages the use of search algorithms for industrial robotic applications, opens up to the study of other robot symmetries present in different robot models and lays a foundation for the application of dynamic obstacle avoidance algorithms.
“…There are also hybrid versions such a cell-based Voronoi roadmap generation algorithm that is searched with A* [10]. Attempts have been made with reinforcement learning [29,30]. All of these algorithms are in most of the cases based on high-dimensional configuration spaces, corresponding to the robot joint positions.…”
This paper argues that an efficient artificial intelligence control algorithm needs the built-in symmetries of an industrial robot manipulator to be further characterized and exploited. The product of this enhancement is a four-dimensional (4D) discrete cylindrical grid space that can directly replace complex robot models. A* is chosen for its wide use among such algorithms to study the advantages and disadvantages of steering the robot manipulator within the 4D cylindrical discrete grid. The study shows that this approach makes it possible to control a robot without any specific knowledge of the robot kinematic and dynamic models at planning and execution time. In fact, the robot joint positions for each grid cell are pre-calculated and stored as knowledge, then quickly retrieved by the pathfinding algorithm when needed. The 4D cylindrical discrete space has both the advantages of the configuration space and the three-dimensional Cartesian workspace of the robot. Since path optimization is the core of any search algorithms, including A*, the 4D cylindrical grid provides for a search space that can embed further knowledge in form of cell properties, including the presence of obstacles and volumetric occupancy of the entire industrial robot body for obstacle avoidance applications. The main trade-off is between a limited capacity for pre-computed grid knowledge and the path search speed. This innovative approach encourages the use of search algorithms for industrial robotic applications, opens up to the study of other robot symmetries present in different robot models and lays a foundation for the application of dynamic obstacle avoidance algorithms.
“…Learning based controllers 36,37 offer a promising solution to address the control challenge. In particular, we take inspiration from a previous work 38 where each actuator is considered an autonomous agent that resides within and shares an environment forming a distributed multi-agent system (MAS). 39 The underlying objective is to enable the agents to coordinate their actions to learn a joint optimum behaviour.…”
Manipulators based on soft robotic technologies exhibit compliance and dexterity which ensures safe human-robot interaction. This article is a novel attempt at exploiting these desirable properties to develop a manipulator for an assistive application, in particular, a shower arm to assist the elderly in the bathing task. The overall vision for the soft manipulator is to concatenate three modules in a serial manner such that (i) the proximal segment is made up of cable-based actuation to compensate for gravitational effects and (ii) the central and distal segments are made up of hybrid actuation to autonomously reach delicate body parts to perform the main tasks related to bathing. The role of the latter modules is crucial to the application of the system in the bathing task; however, it is a nontrivial challenge to develop a robust and controllable hybrid actuated system with advanced manipulation capabilities and hence, the focus of this article. We first introduce our design and experimentally characterize its functionalities, which include elongation, shortening, omnidirectional bending. Next, we propose a control concept capable of solving the inverse kinetics problem using multiagent reinforcement learning to exploit these functionalities despite high dimensionality and redundancy. We demonstrate the effectiveness of the design and control of this module by demonstrating an open-loop task space control where it successfully moves through an asymmetric 3-D trajectory sampled at 12 points with an average reaching accuracy of 0.79 cm + 0.18 cm. Our quantitative experimental results present a promising step toward the development of the soft manipulator eventually contributing to the advancement of soft robotics.
“…A different approach is to rely on neural networks [6] or reinforcement learning [7] for data-driven modeling of the soft robot. In [6], a dynamic model of a soft robot is learned through supervised learning using an auto-regressive network, and is employed for closedloop control by model-based reinforcement learning.…”
Section: Introductionmentioning
confidence: 99%
“…In [6], a dynamic model of a soft robot is learned through supervised learning using an auto-regressive network, and is employed for closedloop control by model-based reinforcement learning. In [7], a multiagent reinforcement learning approach is used to learn the kinematic model of a robotic arm. A trajectory optimization method is also exploited for open-loop control of dynamic reaching tasks [8].…”
Tip control is a current open issue in soft robotics; therefore, it has received a good amount of attention in recent years. The desirable soft characteristics of these robots turn a well-solved problem in classic robotics, like the end-effector kinematics and dynamics, into a challenging problem. The high redundancy condition of these robots hinders classical solutions, resulting in controllers with very high computational costs. In this paper, a simplification is proposed in the actuation setup of the I-Support soft robot, allowing the use of simple strategies for tip inclination control. In order to verify the proposed approach, inclination step input and trajectory-tracking experiments were performed on a single module of the I-Support robot, resulting in zero output error in all cases, including those where the system was exposed to disturbances. The comparative results of the proposed controllers, a proportional integral derivative (PID) and a fractional order robust (FOPI) controller, validate the feasibility of the proposed approach, showing a clear advantage in the use of the fractional robust controller for the tip inclination control of the I-Support robot compared to the integer order controller.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.