All-Weather, sub-50-cm, Radar-Inertial Positioning

Abstract

Deploying automated ground vehicles beyond the confines of sunny and dry climes will require sub-lane-level positioning techniques that use radio waves, rather than near-visible light radiation. Like human sight, LiDAR and optical cameras perform poorly in low-visibility conditions. We present and demonstrate a novel technique for robust, sub-50-cm, urban ground vehicle positioning based on all-weather sensors. The technique incorporates a computationally-efficient, globally-optimal radar scan registration algorithm within a larger estimation pipeline that fuses data from commercially-available, low-cost, automotive radars, low-cost inertial sensors, vehicle motion constraints, and, when available, precise GNSS measurements. We evaluate the performance of the presented technique on an extensive and realistic urban dataset derived from all-weather sensors. Comparison against ground truth shows that during 60 min of GNSS-denied driving in the urban center of Austin, TX, the technique maintains 95th-percentile errors below 50 cm in horizontal position and 0.5 in heading.

Low-Cost Inertial Aiding for Deep-Urban Tightly Coupled Multi-Antenna Precise GNSS

“…This paper adopts the VDC scheme of Narula et al (2022) with minor modifications to the NHC sideslip model and ZUPT detection mechanism. Details of the VDC scheme are provided in this paper's supplemental material.…”

“…Much automated ground vehicle (AGV) research focuses on the use of lidar and cameras for navigation, but these sensing modalities often perform poorly in low-illumination conditions or during adverse weather such as heavy fog or snowy white-outs. By contrast, positioning techniques based on radio waves, such as automotive radar or GNSS, are robust to poor weather and lighting conditions (Narula et al, 2022). Recent work has found that fusing measurements from low-cost automotive radars with inertial sensing can provide lane-level accuracy in urban environments (Narula et al, 2022).…”

“…By contrast, positioning techniques based on radio waves, such as automotive radar or GNSS, are robust to poor weather and lighting conditions (Narula et al, 2022). Recent work has found that fusing measurements from low-cost automotive radars with inertial sensing can provide lane-level accuracy in urban environments (Narula et al, 2022). But radar-based positioning in a global coordinate frame requires the production and maintenance of radar maps, which is a time-consuming and costly endeavor.…”

Micro-Drone Ego-Velocity and Height Estimation in GPS-Denied Environments Using an FMCW MIMO Radar

“…These solutions are thus not adapted to our use-case of the pipings mapping with a microdrone, which requires a low computational cost in order to maximize battery life and the ability to work in a potentially visually-obstructed environment. Another type of sensor that has recently gained attention in the literature on autonomous navigation is radar, because of its high accuracy and robustness to visually degraded conditions [15]. The main disadvantages of radar are its high power consumption and bulkiness which makes it difficult to embark on a micro-drone.…”