We have had success using the navX IMU for sensing robot orientation. We wanted to move to the next level by tracking position with the IMU. We are aware that there are drift issues. However, after the software team reported that the unit was supplying suspect X and Y displacement values, I connected the unit to a Pc and collected data with the navXUI utility. Before collecting data, I ensured that the unit went through a calibration cycle before starting the data collection. I walked round a distance of over 50 ft in different directions, but the data is unsatisfactory. In the attached file I have plotted the X and Y displacement vs the time stamp. I feel that I missing something obvious because other folks seem to be able to use this unit to follow the position of their robot. This mentor is stumped. Any pointers would be most appreciated. The fw has been upgraded to 3.0.34]
navXData_02_04_2017_5_24_53.xlsx (852 KB)
navXData_02_04_2017_5_24_53.xlsx (852 KB)
Please review the displacement FAQ note at Frequently-asked Questions | navX-MXP.
Other teams are typically using other sensors (sometimes including navX-MXP) fused together to get reasonably accurate dead reckoning. These include quadrature encoders (either on drive wheels or ‘follower’ wheels), vision processors, and every now and then LIDAR.