Localization Requirements for Autonomous Vehicles (1906.01061v1)

Abstract: Autonomous vehicles require precise knowledge of their position and orientation in all weather and traffic conditions for path planning, perception, control, and general safe operation. Here we derive these requirements for autonomous vehicles based on first principles. We begin with the safety integrity level, defining the allowable probability of failure per hour of operation based on desired improvements on road safety today. This draws comparisons with the localization integrity levels required in aviation and rail where similar numbers are derived at 10^-8 probability of failure per hour of operation. We then define the geometry of the problem, where the aim is to maintain knowledge that the vehicle is within its lane and to determine what road level it is on. Longitudinal, lateral, and vertical localization error bounds (alert limits) and 95% accuracy requirements are derived based on US road geometry standards (lane width, curvature, and vertical clearance) and allowable vehicle dimensions. For passenger vehicles operating on freeway roads, the result is a required lateral error bound of 0.57 m (0.20 m, 95%), a longitudinal bound of 1.40 m (0.48 m, 95%), a vertical bound of 1.30 m (0.43 m, 95%), and an attitude bound in each direction of 1.50 deg (0.51 deg, 95%). On local streets, the road geometry makes requirements more stringent where lateral and longitudinal error bounds of 0.29 m (0.10 m, 95%) are needed with an orientation requirement of 0.50 deg (0.17 deg, 95%).

• The paper derives strict localization error bounds from road geometry and vehicle dimensions, targeting a SIL of 10⁻⁸ failures per hour.
• It quantifies lateral, longitudinal, vertical, and orientation requirements for freeways and local streets based on detailed numerical analyses.
• The study advocates a multimodal sensor fusion strategy, integrating LiDAR, radar, cameras, GNSS, and IMUs to achieve robust, high-integrity localization.

Localization Requirements for Autonomous Vehicles

The paper "Localization Requirements for Autonomous Vehicles" addresses the critical requirements for precise localization necessary to ensure the safe and efficient operation of autonomous vehicles (AVs). These requirements are derived from fundamental principles and focus on the critical need for accurate positioning and orientation in diverse environmental and traffic conditions. The authors, affiliated with Ford Motor Company, extend methodologies traditionally used in aviation and rail industries to the automotive context.

The primary focus of the paper is the determination of localization requirements that ensure an AV can maintain precise knowledge of its position and orientation within its lane. This is crucial for path planning, perception, control, and general safe operation. The paper begins by establishing a Safety Integrity Level (SIL), analogous to the aviation and rail sectors, which allows a failure probability of 10\textsuperscript{-8} per hour of operation. The authors proceed to derive these requirements from road geometry constraints and vehicle dimensions.

Numerical Results and Key Findings

The paper presents specific numerical requirements for localization. For passenger vehicles operating on freeways, the necessary lateral error bound is 0.57 meters, with 0.20 meters required for 95% accuracy. Correspondingly, longitudinal and vertical bounds are derived as 1.40 meters and 1.30 meters, respectively, with each requiring 95% accuracy thresholds of 0.48 meters and 0.43 meters. The required attitude or orientation bounds are also specified as 1.50 degrees, with a 95% accuracy of 0.51 degrees. In contrast, for local streets with more stringent geometry, the requirements are tighter: lateral and longitudinal bounds of 0.29 meters (0.10 meters for 95% accuracy) are stipulated alongside an orientation requirement of 0.50 degrees (0.17 degrees for 95% accuracy).

Implications and Future Directions

This research has crucial implications for the design and implementation of localization systems in AVs. The stringent accuracy and integrity requirements dictate that current technologies need to evolve or be intelligently combined to achieve these benchmarks consistently under varying conditions. The paper notes that no single localization technology currently meets these requirements independently. Thus, a multi-modal sensor fusion strategy involving LiDAR, radar, cameras, GNSS, and IMUs is essential for high-integrity localization.

The requirements derived in this paper not only lay out a roadmap for technological development and innovation in AV localization systems but also hint at the need for collaboration across industries and between technology providers. The authors suggest that the realization of these requirements might involve trade-offs between onboard sensor fidelity and supporting infrastructure such as high definition maps and GNSS corrections.

In terms of future developments, the paper implicitly encourages exploration into new sensor technologies, resilient algorithms, and robust data fusion methodologies. This is in line with achieving cost-effective and scalable localization solutions that comply with the presented safety and accuracy demands.

Conclusion

The localization requirements presented in the paper are integral to the advancement of autonomous vehicle technology. They present significant challenges, particularly given the demands for high precision and reliability akin to systems used in commercial aviation. Achieving these requirements will necessitate advancements in sensor technology and data processing, paving the way for safer and more efficient autonomous systems. This paper serves as a foundational piece for ongoing research and development efforts in the domain of autonomous vehicle technology.