Recent advances in Rapidly-exploring random tree: A review

Path planning is an crucial research area in robotics. Compared to other path planning algorithms, the Rapidly-exploring Random Tree (RRT) algorithm possesses both search and random sampling properties, and thus has more potential to generate high-quality paths that can balance the global optimum and local optimum. This paper reviews the research on RRT-based improved algorithms from 2021 to 2023, including theoretical improvements and application implementations. At the theoretical level, branching strategy improvement, sampling strategy improvement, post-processing improvement, and model-driven RRT are highlighted, at the application level, application scenarios of RRT under welding robots, assembly robots, search and rescue robots, surgical robots, free-floating space robots, and inspection robots are detailed, and finally, many challenges faced by RRT at both the theoretical and application levels are summarized. This review suggests that although RRT-based improved algorithms has advantages in large-scale scenarios, real-time performance, and uncertain environments, and some strategies that are difficult to be quantitatively described can be designed based on model-driven RRT, RRT-based improved algorithms still suffer from the problems of difficult to design the hyper-parameters and weak generalization, and in the practical application level, the reliability and accuracy of the hardware such as controllers, actuators, sensors, communication, power supply and data acquisition efficiency all pose challenges to the long-term stability of RRT in large-scale unstructured scenarios. As a part of autonomous robots, the upper limit of RRT path planning performance also depends on the robot localization and scene modeling performance, and there are still architectural and strategic choices in multi-robot collaboration, in addition to the ethics and morality that has to be faced. To address the above issues, I believe that multi-type robot collaboration, human-robot collaboration, real-time path planning, self-tuning of hyper-parameters, task- or application-scene oriented algorithms and hardware design, and path planning in highly dynamic environments are future trends.

An improved path planning algorithm based on artificial potential field and primal-dual neural network for surgical robot

Safety and accuracy are essential for path planning in a surgical navigation system. In this paper, an improved path planning algorithm is proposed to increase the autonomous level of spine surgery robots for higher safety and accuracy. Firstly, the dynamic gravitational constant and piecewise repulsion function are adopted to improve the traditional Artificial Potential Field algorithm to solve the common issues of path planning, including local minimum, unable to reach the target near obstacles. To better control the pose of the end-effector in an operation space, the positions of the two endpoints of the end-effector are further constrained. Secondly, an improved Primal-Dual Neural Network with multiple constraints is proposed to minimize the joint angular velocity norm. The multiple constraints are formulated according to the planned path, the obstacle avoidance of the robot and the joint limits. Moreover, a real-time planned velocity scheme is applied to prevent the accumulation of position errors. The simulation results of the pedicle screw implantation demonstrate that the robot can find the collision-free trajectory and arrive at the target position in various complicated situations. More specifically, the error between two endpoints of the end-effector and the target pose is below 0.1 mm in reaching the surgical tool pose, while the maximum position error is around 0.05 mm when performing the planned path. Moreover, two experiments are conducted in the real-world to verify the proposed algorithm is effective in practice.