INERTIAL 2019: Navigine and AD’s brand-new scientific research

We are going to Florida, to the exclusive international Symposium on Inertial Sensors and Systems. The event continues an annual tradition of informal single-track international meetings discussing the latest developments in the area of modern inertial sensors and emerging applications. The INERTIAL 2019 will be a five-day event with one day of tutorials, and four days of technical sessions.

Together with our colleagues from Analog Devices we participate in one of the technical sessions on Wednesday, 3 April. We are going to present the results of our common research «Sensor Fusion for Land Vehicle Localization Using Inertial MEMS and Odometry».

A partnership with Analog Devices had commenced in autumn 2018, and in January 2019 we finished our first common project — arrange of brand-new autonomous cars. The presentation of these drones took place during Consumer Electronics Show (CES) January, 9–12 in Las Vegas, USA. Our car project was not only a cutting-edge marketing tool but also a significant contribution to our further scientific research.

Our study «Sensor Fusion for Land Vehicle Localization Using Inertial MEMS and Odometry» focuses on analysis of current trustworthy self-localization methods and aims at identification of the most appropriate real-time sensor fusion module. Unlike the common approach (to fuse all available information sources: GNSS, IMU equipped with the tri-axial gyroscope and accelerometer, odometer, and perception sensors), IMUs allow for navigation completely independent of external references (e.g. satellites, road markers, geomaps, databases) and immune to weather conditions. In our work, we demonstrate where the gyroscope and accelerometer errors are aided with odometer measurements, GNSS data, and non-holonomic motion constraint. The proposed approach is cost-effective as it requires only a tactical-grade commercial MEMS IMU paired with an OBD-II adapter for odometer measurements and GNSS receiver (when signal is available). The mean circular error after numerous runs without GNSS corrections was 30 meters after 15 minutes (or 7 cm after 30 seconds), which is a solid result for the off-the-shelf MEMS IMU with 0.9 °/hr Bias Instability and 0.2 °/ hr Angle Random Walk.

See you at INERTIAL 2019!