Safe Position Bounds and Fault Detection and Exclusion for Autonomous Driving
Ilaria Martini, Olivier Julien, Chris Hide, Hayden Dorahy, Mehran Khaghani, M. Limberger, u-blox; Ian Sheret, Polymath Insight Ltd.
Date/Time: Wednesday, Sep. 18, 8:57 a.m.
Autonomous transportation and in particular autonomous driving for Advanced Driver Assistance System (ADAS) require highly accurate and reliable positioning. The six automation levels of vehicles driving functions are described in the SAE International J3016-Standard [1]. The full autonomous levels (from 3 to 5) have demanding requirements with centimeter level position accuracy, Alert Limit in the order of 1-2 m and integrity risk in the order of 10e-8 in any hour of operation.
Integrity concepts and augmentation systems (e.g. ARAIM, SBAS) developed for aviation applications and included in aviation standards like RTCA/EUROCAE Dual Frequency Multi Constellation Minimum Operational Requirements Standard [2,3,4] have been designed to work in conditions significantly different than those of automotive and terrestrial applications. An aircraft in horizontal navigation operations (LPV-200) at an altitude above 200 feet have Alert Limit in the order of tens or even hundreds of meters. It is mostly affected by constellation and satellite faults, while multipath is dominated by local reflections of aircraft wings and fuselage. These effects can be characterized a priori and conservatively included in the protection levels [3, 4, 8]. On the other side, terrestrial applications are dominated by local threats like highly correlated multipath, interference, No Line of Sight and signal masking leading to position error distributions with no gaussian tails and highly correlated errors (both over time and among measurements) [5]. Besides, commitments from Constellation Service Providers on satellite and constellation fault probabilities and conservative overbounding techniques provide integrity parameters (i.e. fault probabilities) significantly large to ensure bounds below few meters [6, 7]. Receiver front-end processing might not be aligned to assumptions of augmentation system like Satellite Based Augmentation System leading to additional error sources (e.g. signal distortions). In this context, PPP correction services and sensor fusions with IMU, Wheel Speed Sensors, LIDARs and cameras are key and essential elements to reduce the effects of constellation and atmospheric contributions, provide tight bounds and improve availability in harsh environments.
Standards for PPP corrections and integrity messages are under development in RTCM SC104, SC134 and 3GPP LPP groups. The integrity messages, format and definitions need to be optimized and aligned to the assumptions and models used in the receiver. The selection of the messages and of the Correction Service are crucial to optimize the allocation of the integrity risk among the system elements and simultaneously optimize availability and service coverage.
Another aspect in the safe product development at u-blox concerns the compliance to safety standards. The first edition of the ISO 21448 standard has been published in mid-2022 complementing the well-established Functional Safety (FuSa) standard ISO 26262. The ISO 21448 provides an argument framework and guidance to ensure Safety Of The Intended Functionality (SOTIF), i.e., absence of unreasonable risk due to hazards caused by the system’s functional insufficiencies and thus is considered highly relevant in this context.
SOTIF is about specifying and designing a system that copes with functional insufficiencies, e.g., by dedicated design measures, as well as the development of a SOTIF verification and validation (V&V) strategy. A process is defined at u-blox to structure all SOTIF related activities with documentation via SOTIF work products covering the project lifecycle along the Aspice V-model.
U-blox works on several solutions for autonomous driving and safety critical applications. Previous publications presented concepts based on Bayesian approach to estimate the probability density function of position error and extract Single Epoch Position Bound (SEPB). This approach is based on Student-t distributions fitting posterior error statistic, on single epoch processing to simplify the handling of correlation time, on carrier phase measurements without ambiguity fixing and on delta phase propagation [9, 10, 11]. This paper provides an overview of the updates on the algorithms, on the error model and on the processing optimization which improve the overall performance in different conditions. It describes the process for selecting the measurements and for excluding faults before the bound computation. Optimization is important to reach sufficient and satisfactory level of availability especially in harsh environments where signal masking can drastically reduce the number of available signals. Several techniques are described and used to optimize the selection process and to reduce the conservatism in excluding single measurement. Different criteria are used and combined to ensure that rare outliers are excluded and satisfy the integrity requirements. Sensors are combined to identify and select the reference solution from a safety point of view and to compensate the lack of compliance of single components to availability and safety requirements. In addition, accuracy and prior assumptions on the error model are considered in the selection process to ensure the optimal tradeoff in satisfying performance requirements and to guarantee the consistency with the integrity risk allocation to each system component. The concept and the algorithm performance will be characterized through several hours of real measurement campaigns and the level of integrity and availability performance reached by the solution adopted will be presented and described.
References
[1] SAE International J3016-Standard: Taxonomy and Definitions for Terms Related to Driving Automation Systems for On-Road Motor Vehicles available at https://www.sae.org/standards/content/j3016_202104/
[2] EUROCAE, “ED-259A – Minimum Operational Performance Standard for Dual-Frequency Multi-Constellation Satellite based Augmentation System Airborne Equipment.” France, 2023.
[3] Blanch, J., Walter, T., Enge, P., Lee, Y., Pervan, B., Rippl, M., Spletter, A., Kropp, V., "Baseline Advanced RAIM User Algorithm and Possible Improvements," IEEE Transactions on Aerospace and Electronic Systems, Volume 51, No. 1, January 2015.
[4] Working Group C, ARAIM Technical Subgroup, Milestone 3 Report, February 26, 2016. Available at: http://www.gps.gov/policy/cooperation/europe/2016/working-group-c/
http://ec.europa.eu/growth/tools-databases/newsroom/cf/itemdetail.cfm?item_id=8690
[5] Langel, Steven, Crespillo, Omar García, Joerger, Mathieu, "A New Approach for Modeling Correlated Gaussian Errors Using Frequency Domain Overbounding," 2020 IEEE/ION Position, Location and Navigation Symposium (PLANS), Portland, Oregon, April 2020, pp. 868-876.
[6] T. Walter, J. Blanch, K. Gunning, M. Joerger and B. Pervan, "Determination of Fault Probabilities for ARAIM," in IEEE Transactions on Aerospace and Electronic Systems, vol. 55, no. 6, pp. 3505-3516, Dec. 2019, doi: 10.1109/TAES.2019.2909727. keywords: {Global Positioning System;Satellite broadcasting;Satellites;FAA;Distance measurement;Noise measurement},
[7] Perea, S., Wallner, S., Schönfeldt, M., Binder, K., Odriozola, M., Donatelli, A., Foucault, E., Sgammini, C. Stallo M., Martini, I., Boyero, J.P., Mabilleau, M., Canestri, E., Castrillo, N., "Galileo H-ARAIM: Update on Performance Characterization and Integrity Support Message," Proceedings of the 35th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2022), Denver, Colorado, September 2022, pp. 683-701. https://doi.org/10.33012/2022.18404
[8] Circiu, Mihaela-Simona, Caizzone, Stefano, Enneking, Christoph, Fohlmeister, Friederike, Rippl, Markus, Meurer, Michael, Felux, Michael, Gulie, Ioana, Rüegg, David, Griggs, Joseph, Lazzerini, Rémy, Hagemann, Florent, Tranchet, Francois, Bouniol, Pierre, Sgammini, Matteo, "Final Results on Airborne Multipath Models for Dualconstellation Dual-frequency Aviation Applications," Proceedings of the 2021 International Technical Meeting of The Institute of Navigation, January 2021, pp. 714-727. https://doi.org/10.33012/2021.17862
[9] Bryant, Rod, Julien, Olivier, Hide, Chris, Skorupa, M., Dorahy, H., Sheret, Ian, "Safety-Critical Automotive Positioning Based on SEPB without Atmospheric Corrections," Proceedings of the 34th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2021), St. Louis, Missouri, September 2021, pp. 1843-1858.
[10] Bryant, Rod, Julien, Olivier, Hide, Chris, Skorupa, M., Sheret, Ian, "Road Vehicle Integrity Bound Propagation Using GNSS/IMU/Odometer," Proceedings of the 33rd International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2020), September 2020, pp. 585-611.
[11] Julien, Olivier, Hide, Chris, Dorahy, Hayden Hide, Chris, u-blox; Ian Sheret, Polymath Insight Ltd “Extended Results of Single Epoch Position Bound (SEPB) for High Integrity Automotive Applications” Proceedings of the 37rd International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2023), September 2023
For Attendees Call for Abstracts Registration Hotel Travel and Visas Exhibits Submit Kepler Nomination For Authors and Chairs Abstract Management Student Paper Awards Editorial Review Policies Publication Ethics Policies For Exhibitors Exhibitor Resource Center Marketing Resources Other Years Future Meetings Past Meetings