Abstract:Purpose This paper aims to investigate an autonomous obstacle-surmounting method based on a hybrid gait for the problem of crossing low-height obstacles autonomously by a six wheel-legged robot. The autonomy of obstacle-surmounting is reflected in obstacle recognition based on multi-frame point cloud fusion. Design/methodology/approach In this paper, first, for the problem that the lidar on the robot cannot scan the point cloud of low-height obstacles, the lidar is driven to rotate by a 2D turntable to obtain the point cloud of low-height obstacles under the robot. Tightly-coupled Lidar Inertial Odometry via Smoothing and Mapping algorithm, fast ground segmentation algorithm and Euclidean clustering algorithm are used to recognize the point cloud of low-height obstacles and obtain low-height obstacle in-formation. Then, combined with the structural characteristics of the robot, the obstacle-surmounting action planning is carried out for two types of obstacle scenes. A segmented approach is used for action planning. Gait units are designed to describe each segment of the action. A gait matrix is used to describe the overall action. The paper also analyzes the stability and surmounting capability of the robot’s key pose and determines the robot’s surmounting capability and the value scheme of the surmounting control variables. Findings The experimental verification is carried out on the robot laboratory platform (BIT-6NAZA). The obstacle recognition method can accurately detect low-height obstacles. The robot can maintain a smooth posture to cross low-height obstacles, which verifies the feasibility of the adaptive obstacle-surmounting method. Originality/value The study can provide the theory and engineering foundation for the environmental perception of the unmanned platform. It provides environmental information to support follow-up work, for example, on the planning of obstacles and obstacles.

Identifying and Approaching for Obscured Stairs

An Environment Recognition Algorithm for Staircase Climbing Robots

A Method to Detect Stairs with Three-Dimensional Scanning for Hexapod Robot Stair Climbing

ATV Navigation in Complex and Unstructured Environment Containing Stairs

Staircase Localization for Autonomous Exploration in Urban Environments

RGB-D-based Stair Detection using Deep Learning for Autonomous Stair Climbing

Deep leaning-based ultra-fast stair detection

An designated obstacle monitoring approach based on self-defined landmarks for a mobile robot

Safe and Robust Mobile Robot Navigation in Uneven Indoor Environments

An Onboard Framework for Staircases Modeling Based on Point Clouds

A Robust Vision-Based Method for Staircase Detection and Localization

Lidar-Artificial-Marker Odometry for a Surface Climbing Robot Via Factor Graph

A Review of the Application of Staircase Scene Recognition System in Assisted Motion

Obstacle Detection and Obstacle-Surmounting Planning for a Wheel-Legged Robot Based on Lidar

StairNetV3: Depth-aware Stair Modeling using Deep Learning

Obstacle Detection and Avoidance Methods for Mobile Robot in Indoor Environment

Design and Implementation of Autonomous Navigation System Based on Tracked Mobile Robot.

Vision Only Autonomous Indoor Exploration Method for Unmanned Aerial Vehicles (Uavs)

Autonomous Indoor Exploration Of Mobile Robots Based On Door-Guidance And Improved Dynamic Window Approach

A Real-Time Multi-Constraints Obstacle Avoidance Method Using LiDAR

Real-Time Avoidance Strategy of Dynamic Obstacles via Half Model-Free Detection and Tracking With 2D Lidar for Mobile Robots