Robotics and Sensor-Based Localization

Robots and autonomous vehicles need to answer two questions simultaneously: where am I, and what does the space around me look like? Simultaneous Localization and Mapping (SLAM) addresses this by fusing sensor data — from cameras, lidar, or depth sensors — into a consistent geometric model of the environment while continuously estimating the robot's position within it. Techniques like visual odometry track motion by analyzing sequential image frames, and point cloud processing turns raw depth measurements into navigable 3D maps, with RGB-D cameras and monocular setups representing different trade-offs between cost, weight, and accuracy. Active research focuses on making these systems robust under difficult conditions — low light, featureless terrain, rapid motion — and on scaling accurate real-time mapping to the large, dynamic environments encountered in aerospace applications such as planetary rovers and autonomous aerial vehicles.

Works

92,685

Total citations

1,393,595

Keywords

SLAM3D MappingVisual OdometryRoboticsLocalizationPoint Cloud