Abstract: | Recently, autonomous driving technology has being matured. Google is leading the development race, their self-driving car has already ran more than 500,000 km on public roadway with No accident and they have announced to unveil their first complete prototype self-driving car. Most of car manufacturers are following and accelerating autonomous driving technology. The current autonomous driving technology relies on several sensors and three dimensional digital map. GNSS is one of the sensors to measure absolute position of vehicle, however, its importance, priority on the localization algorithm is relatively lower than other sensors like vision sensor, millimeter wave radar, IMU and laser profiler. Car runs many different environments which are severe conditions for continuous satellite signal reception, where there are many multi-path, undirect signals, such as tunnel, urban canyon, under canopy, bridge and so on. However, due to progress on the dployment of multiple GNSS constellations, the increasing visible satellite can contribute to the improvement of Positioning, Navigation and Timing (PNT) performance even in the such severe conditions. Japanese government, Cabinet Office, launched a R&D scheme in 2014 called “Strategic Innovation Promotion Program (SIP)”. The scheme is aiming to promoting innovative R&D activities in cross-ministerial and integrated manner, while each ministry conducted independently past. Ten challenging subjects were selected including autonomous driving technology. Under this SIP framework, it was proposed that the effect of multiple GNSS constellation use on the self-driving and/or advanced driving assist would be assessed by using collected data in several different environments. A consortium lead by AISAN Technology Co. Ltd, including JAXA, Satellite Positioning Applications promotion Center (SPAC), Electronic Navigation Research Institute (ENRI) and Tokyo University of Marine Science and Technology (TUMSAT) conducted test driving, analysis and assessment. Three days test driving was conducted on December 13, 14 and 15, 2014. Driving route was planned so as to cover all environments car experienced from open sky on the highway to dense urban canyon surrounding skyscraper in Tokyo downtown. Japanese car manufacturer requested to include tunnel, multi-layer road which covered by other road and elevated railway which run in parallel, crossing overhead pedestrian way, and highway with tall sound insulating walls. Test vehicle has Mobile Mapping System (MMS) as reference, several types of geodetic grade receivers, single frequency code phase positioning receivers with different antenna locations were on board. In addition, two RF recoders were also installed and RF data were recorded through the test driving for future study that receiver”s algorithm would be investigated further. GPS, GLONASS, BeiDou and QZSS were observed and post processing analysis compared with regard to different combination of GNSS as well as several augmentation methods. The collected data including RF data will be opend to the public after analysis team will conduct the assessment and report the result. The assessment result is described in the paper with items to be resolved for the GNSS applications on car self-driving and advanced driver assistance technology. |
Published in: |
Proceedings of the 28th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2015) September 14 - 18, 2015 Tampa Convention Center Tampa, Florida |
Pages: | 495 - 512 |
Cite this article: | Kogure, Satoshi, Hosoi, Mikihiro, Kubo, Nobuaki, Sakai, Takeyasu, Matsuoka, Shigeru, "Assessment of the Benefit of Multiple GNSS Constellation use in Tokyo Downtown for Future Autonomous Driving and Advanced Driver Assistance System," Proceedings of the 28th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2015), Tampa, Florida, September 2015, pp. 495-512. |
Full Paper: |
ION Members/Non-Members: 1 Download Credit
Sign In |