Premium
Localizing Ground Penetrating RADAR: A Step Toward Robust Autonomous Ground Vehicle Localization
Author(s) -
Cornick Matthew,
Koechling Jeffrey,
Stanley Byron,
Zhang Beijia
Publication year - 2016
Publication title -
journal of field robotics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.152
H-Index - 96
eISSN - 1556-4967
pISSN - 1556-4959
DOI - 10.1002/rob.21605
Subject(s) - global positioning system , lidar , inertial measurement unit , computer science , real time computing , ground truth , remote sensing , multipath propagation , radar , gps signals , assisted gps , artificial intelligence , simulation , geography , telecommunications , channel (broadcasting)
Autonomous ground vehicles navigating on road networks require robust and accurate localization over long‐term operation and in a wide range of adverse weather and environmental conditions. GPS/INS (inertial navigation system) solutions, which are insufficient alone to maintain a vehicle within a lane, can fail because of significant radio frequency noise or jamming, tall buildings, trees, and other blockage or multipath scenarios. LIDAR and camera map‐based vehicle localization can fail when optical features become obscured, such as with snow or dust, or with changes to gravel or dirt road surfaces. Localizing ground penetrating radar (LGPR) is a new mode of a priori map‐based vehicle localization designed to complement existing approaches with a low sensitivity to failure modes of LIDAR, camera, and GPS/INS sensors due to its low‐frequency RF energy, which couples deep into the ground. Most subsurface features detected are inherently stable over time. Significant research, discussed herein, remains to prove general utility. We have developed a novel low‐profile ultra‐low power LGPR system and demonstrated real‐time operation underneath a passenger vehicle. A correlation maximizing optimization technique was developed to allow real‐time localization at 126 Hz. Here we present the detailed design and results from highway testing, which uses a simple heuristic for fusing LGPR estimates with a GPS/INS system. Cross‐track localization accuracies of 4.3 cm RMS relative to a “truth” RTK GPS/INS unit at speeds up to 100 km/h (60 mph) are demonstrated. These results, if generalizable, introduce a widely scalable real‐time localization method with cross‐track accuracy as good as or better than current localization methods.