Hello,

Recently I tested my robot running gmapping/karto/cartographer in a building with glass doors and glass walls, and unfortunately laser does not reflect from the glass, so performance was no good. Also, the laser beam reflects some of the time from the glass when the angle of incidence is a certain degree. I guess at 0 degrees of incidence, the laser does reflect back and also I am suspecting the phenomenon known as total internal reflection causes the lidar to get different readings at different degrees of incidence.

This further confuses the mapping algorithms.

How can we overcome this problem? Right now, I have a YDLidar X4. Could I get away by buying a more expensive LIDAR that has more precision in reading distances? X4 is a triangulation based scanner. The same vendor also produces ToF based models. Could a ToF based model perform better in a environment with glass doors and walls?

Are there any other solutions where we can augment the lidar system with lets say ultrasonic rangefinders. If so could we integrate this to the SLAM/Navigation system?

Best Regards, C.A.