3 January 2024 in International Journal of Control, Automation, and Systems
16 March 2022 in Intelligent Service Robotics
This research introduces a new path planning method for rescue robots in a dynamic and partially known area when the robots are performing tasks in a large area. The path planning of the rescue robots that move in the dynamic area is introduced to solve the issue of unnecessary areas, which are the disadvantages of the existing D*-based algorithms. This paper proposes a method to eliminate unnecessary expanded nodes of the dynamic and partially known environment by segmenting a map with an auto-clustering algorithm, which is able to achieve a faster execution time than conventional algorithms. Furthermore, to show the effectiveness of the proposed algorithms, an expected value of re-planned nodes in the dynamic and partially known area is introduced using a probability-based approach.
8 January 2024 in IEEE Robotics and Automation Letters
Traditional sampling-based algorithms rely on random samples to explore a whole configuration space of robots for optimal path planning, while a uniform sampler impedes the exploration with randomly generated samples, leading to a long calculation time, especially in complex environments. Recently, neural-network-driven methods have attracted wide interest in developing non-uniform sampling to improve the sampling efficiency and reduce the calculation time. A region that contains an optimal path is predicted by neural networks and employed subsequently to biasedly generate samples. This work aims at enhancing the sampling efficiency and reducing the calculation time of the optimal path planning by a novel region and guideline prediction (denoted as RGP) model. We innovatively propose the RGP model with a guideline prediction module to estimate the guideline distributions, which are characterized by the central line of the predicted region. The predicted region and guideline are integrated into a sampling-based algorithm, namely RGP-RRT*, with an adaptively biased sampling strategy to select a proper domain for sampling. Simulations demonstrate the RGP model outperforms other region prediction models in accuracy and robustness. Besides, the RGP-RRT* reliably achieves a 7.2–80.1% reduction in calculation time and a 2.0–58.1% reduction in sample number compared with other neural-network-driven methods.
20 August 2021 in Journal of Robotics and Mechatronics
A tele-rehabilitation system that can achieve remote interaction between a physical therapist (PT) and a patient was developed. Patients need to execute rehabilitation exercises to maintain upper limb function. However, it is difficult for them to travel to hospitals without aid. This system is equipped with a PC and a Kinect sensor at the hospital side (i.e., the PT), and a PC and an upper limb assistive device in the patient’s home. The PT displays the motion in front of a Kinect sensor, which identifies the motion. In addition, the device on the home side assists the motion of the patient using the Internet. When the device receives a force higher than the safety value from the patient at any particular point on it, vibrators at the corresponding point on the PT’s arm start to vibrate. Thereby, the PT can identify the patient’s condition and limitations. The time delays in the transmission of data of device motion and the vibrators were measured and compared. As a result, the PT could identify the patient’s condition faster than the motion of the device.
8 November 2023 in Artificial Life and Robotics