For a mobile robot to move autonomously, the core is autonomous localization and navigation. This involves localization, mapping, and path planning—among which the quality of the map directly affects the robot’s travel path.
Like humans who draw maps to understand and describe an environment, robots rely on maps to represent their surroundings. The map form depends on the chosen algorithms and sensors. In robotics, common representations include occupancy grid maps, feature-based maps, direct (dense) representations, and topological maps. Among these, the occupancy grid is the most widely used in practical applications. Although it looks like a conventional map, it is essentially a bitmap in which each “pixel” encodes the probability that an obstacle exists at that location in the real world.
On the other hand, grid maps can express many environmental features and support path planning; on the other, they do not store raw sensor data directly, achieving good efficiency in space and time. SLAMTEC’s SLAMWARE localization and navigation solution adopts this map type. Comprising the modular autonomous localization and navigation system SLAMWARE Core paired with LiDAR, SLAMWARE—one of the most representative products in laser SLAM—can directly complete map construction, autonomous path planning, and motion control without extra computing, making system integration convenient for users.
In map construction, LiDAR is a crucial tool—it acts as the robot’s “eyes,” sensing the surrounding environment. Before mapping, the LiDAR must be installed correctly; installation accuracy directly affects subsequent localization and mapping. The following uses SLAMTEC’s RPLIDAR as an example.
SLAMTEC LiDAR Installation Methods and Notes
-
Horizontal mounting angle: 0° to +2° (tilt upward).
-
Scan opening angle: Recommended ≥270°; ideally there are openings directly in front and behind the robot; must be >220° at minimum.
-
LiDAR blind zone: ~15 cm. When installing, place the LiDAR slightly closer to the chassis center to reduce the blind zone.
Preparations Before Mapping
-
Perform mapping when the area has few or no people.
-
Remove objects that are frequently moved (e.g., people passing through, movable tables, potted plants).
-
For reflective, transparent, or absorptive objects (e.g., mirrors, glass), if conditions allow, apply other materials at the LiDAR scan plane that do not interfere with scanning.
Mapping Tips
1. At startup, face the robot toward a straight wall, and let the LiDAR capture as many valid points as possible.
2. In malls or other large environments, first guide the robot to close the loop of the map, then scan smaller details thoroughly.
3. Erase unexplored regions: During operation, areas not fully scanned (often limited by LiDAR range) may appear. In SLAMTEC RoboStudio, display these areas (e.g., the red-circled regions) as gray. If there are no obstacles within such regions, use the white eraser to clear the gray, marking them as explored. (See RoboStudio toolbar—the white eraser is highlighted.)
4. Besides removing unexplored areas, you can also erase moving objects in the map (e.g., crowds, obstacles that will be moved later).
5. Try to maximize map completeness: along all potential robot paths, ensure 360° around the robot has been scanned to increase reference points for laser matching.
6. Once the map is built, avoid repeated back-and-forth motion over the completed area. Save the map promptly: click Sync Map in RoboStudio to synchronize and then save.
Checking Mapping Quality
After the map is constructed, evaluate whether it meets localization/navigation requirements using the following criteria:
1. Are obstacle edges in the map clear?
2. Are there regions inconsistent with the real environment (e.g., missing loop closures)?
3. Are there gray areas within the robot’s operational region (i.e., unscanned zones)?
4. Does the map contain obstacles that won’t exist during subsequent localization (e.g., temporary/moving obstacles)?
5. Can the robot, at any position within its activity area, rely on a fully explored 360° field of view?
If you encounter any issues during operation, contact SLAMTEC technical support at support@slamtec.com for professional assistance.
How Much Do You Know About Laser SLAM Mapping Principles and Tips?
Comparison of Laser SLAM and Visual SLAM: Advantages and Disadvantages
