Guinea-pig testing casts a ‘shadow’ over driverless tech

Several US states have already taken the plunge at legalising autonomous vehicle testing and look set to extend this to fully supporting the auto industry’s drive towards mass market adopting of the technology.

However, there are some who have serious reservations at the speed the technology is being introduced and warn that robust regulation must be in place before carmakers are given a free hand to introduce higher levels of autonomy.

Regulation – inhibitor or innovator?

The need for adequate legislation is highlighted by Michael DeKort, former engineering manager at Lockheed Martin, member of SAE’s task force, who is against legalising autonomous technology in its present form. He believes that “there are some issues regarding how autonomous vehicles are being engineered. The use of shadow driving for AI engineering and testing, as well as Level 3, are dangerous and unnecessary”. Shadow driving modes are used by carmakers such as Tesla to test whether their car’s autonomous systems would have worked in multiple scenarios had they been in operation.

He adds: “As the automakers sell cars with the Level 3 capability, and the autonomous companies who use this practice for AI engineering and testing to create autonomous vehicles, move from driving in benign or easy conditions, the accident, injury and death rate will increase. The reason for this is that in order for the AI to learn how to handle every core scenario, including dangerous, complex and actual crash scenarios, they have to be driven. The motive that this is deemed acceptable is because it is believed this process is the best or only way to attain autonomy and that the tragedies along the way are a necessary evil. The practices of shadow driving and Level 3 are dangerous. Their use in the public arena, or for companies to perform the AI engineering and testing to create autonomous vehicles, will create thousands of needless injuries and casualties.”

DeKort emphasises the point that building a self-driving car that meets the reliability requirements equal to our current system is one of the most challenging technological developments ever attempted. He explained: “It requires building a very sophisticated, reliable electromechanical control system with artificial intelligence software that needs to achieve an unprecedented level of reliability at a cost the consumer can afford. Boeing claims a 99.7% reliability figure for its 737 which is equivalent to about 3,000 failures per million opportunities. A modern German Bosch engine control module achieves a reliability of about 10 failures per million units which is about six times worse for a single component than our current system of flawed human drivers. This means that to meet the reliability requirements, all systems and sensors that control any aspect of the car’s movement must be double redundant which really means building nearly two cars in one body!”

The legislative tipping point

It’s not only the vehicle’s transport function that is causing concern among many in the industry. As murmurs of the need for more legislation are getting louder, Yoni Heilbromm, CMO with Argus Cyber Security said: “Recent legislative efforts by the Senate and the House mark a significant step towards finalising autonomous vehicle legislation that mandates multi-layered automotive cyber security as a core requirement. With the reconciliation process underway, it’s possible that final legislation will be approved before the year is out. Lawmakers and vehicle manufacturers alike understand that increased vehicle connectivity and autonomy brings increased risk and these bills recognise the importance of multi-layered automotive cyber security with intrusion detection and prevention systems (IDPS) and of solutions that adjust to unforeseen future circumstances.”

Ritukar Vijay, head of technology and strategy at the Indian company Hi-Tech Robotic Systemz agrees that legislation is a vital step for autonomous vehicles. He said: “2017 started with a large acquisition of Mobile eye by Intel for massive $15Bn (£11Bn). That is a pure indicator that any or all mass adoption of autonomous vehicles (AVs) will be incremental in nature. Regardless the level of technology, companies like Uber, GM are doing self-driving trials with having human driver in the loop with a safety driver in the car. So, when it comes to legislative aspects for AVs, for the next few years it is going to be all about Level 2 to Level 4 autonomy with a human in the loop. However, Level 5 is a clear win for environmentally focused autonomy and its deployment will be seen across the world including India.”

However, Vijay warns: “There is a lot of innovation still to be done when it comes to low cost sensors with high fidelity, be it a single type or multiple sensors fusing together to provide appropriate perception to the AVs.”

NHTSA’s automated vehicle testing and development guidelines

On NHTSA’s automated vehicle testing and development guidelines, Heilbromm has some strong opinions, he says: “Combined with NHTSA’s voluntary guidelines and pending policy regulations, these actions are a clear signal to automakers that have yet to embrace cyber security as a priority that now is the time to be proactive and to take the wheel in adopting effective standards of automotive cyber security. Ensuring the safety and security of drivers and their vehicles is at the core of what we do and we encourage all stakeholders to continue identifying and implementing the best in automotive cyber security standards.”

DeKort feels that: “NHTSA has made the situation worse by producing a flawed report that enables the autonomous driving behaviour. That report is based on incomplete and misleading testing. In addition, the practice of shadow driving to create autonomous vehicles will require so many miles to be driven at such a massive expense (one trillion and over $300Bn (£223Bn) that this process will never result in a fully autonomous vehicle.”

On the same note Vijay feels that: “M-city like infrastructure in various parts of the globe have been setup and it is in sync with NHTSA testing and development guidelines. This is definitely a good step in the recent past but for the coming years there is a great need for standards and certifications which can be mapped vis-a-vis the capabilities of an autonomy service provider or OEMs. This it has to be moved up from just guidelines to make it safer and more legislative friendly.”


Leave a comment

Your email address will not be published. Required fields are marked *