Propagation of Measurement Uncertainty in IMU Orientation Tracking Algorithms

In today's technological scenario, where precision and accuracy are critical, the fusion of data derived from different sensors has become the key to improving measurements in various applications. Among the myriad sensor combinations available, harnessing the power of accelerometer-gyroscope fusion through a complementary filter is a remarkably effective approach. The concept behind the complementary filter is to combine data from an accelerometer and a gyroscope to improve motion detection accuracy and reliability. It uses the low-frequency response of the accelerometer for static orientation and the high-frequency measurements of the gyroscope for dynamic motion detection, resulting in accurate positioning information. Assessing the quality of these methodologies, particularly evaluating the uncertainty of the results obtained, is of great significance. What the authors propose is to assess the uncertainty of the results obtained through the complementary filter by making explicit the functional relationships that allow the angles of interest to be calculated from the accelerometer and gyroscope data and to apply the law of propagation of uncertainty to these as an alternative to the time-consuming and computationally disadvantageous Monte Carlo analysis, making the appropriate considerations on the functional relationships involved.