Application and performance of laser speckle odometry applied to a mobile industrial robot

In this paper we report the application of a laser speckle odometer to a mobile industrial robot in a typical factory floor environment. The suitability of typical floor surfaces and features is assessed in terms of the ability to form speckle patterns with sufficient signal to noise for correlation-based processing. All tested surfaces including concrete, rubber tile, dried paint and oil stains, and hazard tapes were found to be suitable. A comparison of the velocimetry sensor output to the industrial robot’s internal SLAM and wheel encoder data is presented with good agreement of < 0.3mm/s at tested speeds of up to 250mm/s. Finally, a comparison of speckle odometry to the robot’s internal SLAM based navigation will be presented using a laser tracker to provide ground-truth measurement data. Both techniques were found to perform similarly, with errors of up to 80mm when traversing a 16m square path of 4m sides. The laser speckle odometry was however found to perform significantly better over the initial sides of the path with a maximum error of < 10mm in comparison to < 47mm for the robot’s internal odometry.