Safe Navigation in Unknown Dynamic Environments with Voronoi Based StRRT

Abstract

For a robot to move autonomously in an unknown in the workspace. When utilizing this technique in real-time dynamic environment a real-time motion generation is necessary. motion planning, the efficiency of searching for the path can Especially, safety motion generations are impoant. There are a be further improved by partially reusing or refining an existing lot of methods for making safe motion. For example, it is popular to check only around the robot with additional sensors and avoid tree. obstacles detected by the sensors though the coordination problem A common approach to handle moving obstacles utilizes a between global trajectory making and local obstacle avoidance should configurationtime space which consists in adding the time be occrued in this way. It is also popular to estimate the size dimension to the robot's configuration space [7]. Although this of obstacles to be large though adequate width of the dilatation method has usually been applied to motion planning in a is different by environment. Another approach would be to deter-ald to

Bridging Requirements, Planning, and Evaluation: A Review of Social Robot Navigation

“…Our research regarding the applications of classical roadmap methods shows that they are rarely used in social robot navigation as they only consider binary environment models (obstacle or free space); hence, human awareness cannot be properly tackled. However, Voronoi diagrams might be used as reference path-planning approaches [204,[277][278][279] for capturing the skeleton of the environment along with human-aware trajectory planners as in [132].…”

“…Besides global path planning Section 5.1.4, sampling-based methods can also be applied to local trajectory planning. An extended RRT with a notion of time included-spatiotemporal RRT-was proposed by Sakahara et al [204]. Their method integrates ideas of the RRT and the Voronoi diagram.…”

“…In applications where performance is crucial, the constant velocity model, assuming piecewise constant velocity with white noise acceleration, can be applied. Despite its simplicity, it is commonly chosen as an ad hoc method for motion prediction in numerous approaches [139,[203][204][205][206][207][208] having lightweight and straightforward implementation and yielding satisfactory results with high-frequency updates. Recently, Schöller et al [209] discussed that the constant velocity model might outperform state-of-the-art neural methods in some scenarios.…”

Human-centered X–Y–T space path planning for mobile robot in dynamic environments

“…We define constraints of motion based on the maximum acceleration and maximum velocity as the robot's dynamic constraints. Sakahara et al proposed X -Y -T space path planning using rapidly-exploring random trees (RRT), which takes into account the robot's dynamic constraints [18]. High-speed processing of RRT enables real-time operation even if there are multiple obstacles.…”