Sensor Data Fusion for a Mobile Robot Using Neural Networks

Abstract

Mobile robots must be capable to obtain an accurate map of their surroundings to move within it. To detect different materials that might be undetectable to one sensor but not others it is necessary to construct at least a two-sensor fusion scheme. With this, it is possible to generate a 2D occupancy map in which glass obstacles are identified. An artificial neural network is used to fuse data from a tri-sensor (RealSense Stereo camera, 2D 360° LiDAR, and Ultrasonic Sensors) setup capable of detecting glass and other materials typically found in indoor environments that may or may not be visible to traditional 2D LiDAR sensors, hence the expression improved LiDAR. A preprocessing scheme is implemented to filter all the outliers, project a 3D pointcloud to a 2D plane and adjust distance data. With a Neural Network as a data fusion algorithm, we integrate all the information into a single, more accurate distance-to-obstacle reading to finally generate a 2D Occupancy Grid Map (OGM) that considers all sensors information. The Robotis Turtlebot3 Waffle Pi robot is used as the experimental platform to conduct experiments given the different fusion strategies. Test results show that with such a fusion algorithm, it is possible to detect glass and other obstacles with an estimated root-mean-square error (RMSE) of 3 cm with multiple fusion strategies.

Aspects of autonomous drive control using NVIDIA Jetson Nano microcomputer

“…Further research will allow the system to be expanded with additional sensor modules for more precise environment monitoring (including obstacles detection) using data fusion techniques [17].…”

Reinforcement Optimization Algorithm for Mobile Robot Sensor Networks Drive Motion Improvement

“…The four-reinforcement learning method is proposed for the mobile robot kinematic moving model from the moving velocity rate inside the sensor operating regions from the boundary limit. The standardized path planning method considers as a reference from literatures [11]- [13], which are clearly considered to be connecting points from sensor values. Probabilistic path map is created for the random nodes from the starting and end path of the boundary considering the obstacles.…”

“…FOR MOBILE ROBOT SENSOR NETWORKS Introduction of obstacles in the mobile robot path from sensor readings has an impact on linear movement and control movement. Selection of reference from the literatures [11]- [13] bounded values to create a probabilistic path map from the proposed proximal policy optimization, trust region policy optimization, policy gradient reinforcement learning optimization, and deep "Q" network reinforcement learning optimization.…”

“…Navigation architecture has gained attention in recent development of localization algorithms of existing sensor data units; however, path and obstacle clustering must be improved. The collection of more sensor measurement units reduces the computational error, but takes time to identify the localization regardless of the map environment [11]-Manuscript received 17 February, 2022; accepted 11 May, 2022. [13].…”

Multi-LiDAR Mapping for Scene Segmentation in Indoor Environments for Mobile Robots

“…For that reason, it is essential to exploit the uses of environment information coming from diverse sensor sources for achieving a robust indoor representation. Other works have already proven the importance of using diverse sensor sources for overcoming problems such as precise geometric localization [ 1 ] or mapping problematic regions with glass surfaces [ 2 ]. However, previous reviewed papers have stated that sensor fusion using LiDAR and camera sensors is very sensitive to daylight because it interferes with the IR light.…”