11 Companies Weigh In on AV Principles
July 1, 2020—Eleven companies have weighed in on setting electric vehicle principles in order to develop and validate a safe automated driving system.
The goal of the publication, titled "Safety First for Automated Driving," is to summarize safety by design and verification and validation methods of SAE, L3, and L4 automated driving. The companies developing these new AV principles include: Aptiv, Audi, Baidu, BMW, Continental, Daimler, Fiat Chrysler Automobiles, Here, Infineon, Intel, and Volkswagen.
The publication outlines 12 guiding principles of automated driving:
- Safe Operation
- Operational Design Domain
- Vehicle Operator-Initiated Handover
- User Responsibility
- Vehicle-Initiated Handover
- Interdependency Between Vehicle Operators and Automated Driving Systems (ADS)
- Safety Assessment
- Data Recording
- Passive Safety
- Behavior in Traffic
- Safe Layer
In short, the principles aim to minimize risk if the driver does not comply with a takeover request, as well as verification and validation tests intended to ensure safety. The companies feel it is important that the automated vehicles record data, compliant with data privacy laws, of course, and that the vehicles are predictable when it comes to following rules on the road and recognizing system limits.
When it comes to cybersecurity, the companies recommend Secure Development Lifecycle (SDL), a process for installing security that is tailored to fit product development life cycles while considering things like risk treatment strategy, system state, and risk treatment manifestation. The publication also advises that ADS enable localization through sensors, map data, and sensor fusion algorithms so as to prevent autonomous driving in areas where it’s restricted.
The publication goes on to encourage the implementation of the Safety of the Intended Functionality (SOTIF), which avoids unreasonable risks that might happen even if all of a vehicle’s components are operating correctly, such as when the system confuses road signals or traffic signs. The companies believe it will reduce known and unknown potential risks and behaviors.
The companies also suggest that vehicles’ perception sensors should capture all relevant external information with the vehicle's surroundings, like pedestrians, obstacles, traffic signs, and acoustic signals, to minimize risk. These sensors include cameras, microphones, lidar, radar, and ultrasonic.
The strategies highlight what the report refers to as “safety by design,” which would dive into scenario-based automated driving technologies and analyze the systems when they are actually on the road.
According to the report, level 3 and level 4 vehicles face daunting challenges no matter how they are designed. The companies note that tests involving driver interaction will need to take place, and they will need to pass them with flying colors while also proving that the driver is capable of coping with “scenarios not currently known” in traffic.
Overall, the co-authors note that this publication is merely the first draft of their plans, and that the next version of the report will be an official proposal for standardization.