Imagine the following situation: you just received your brand-new, expensive sports car and your friend asks you whether they can take it for a spin. Do you trust your friend to bring back your car without a scratch?
Fast forward: same situation, but a computer offers to drive your car? Do you trust the algorithm? And, do you want that computer driving your car – what about the “joy of driving”?
Fully autonomous driving (SAE level 4 or 5) is on the horizon and creates questions as outlined above which go beyond the technical details of this technology. Car manufacturers face further challenges. For example, in many jurisdictions including many relevant markets such as Europe, China and USA, fully autonomous driving is not allowed in general at the moment. When will these provisions change and how? The answer to these questions will depend on considerations as to what happens if a fully autonomous driving car causes an accident. For example, who is responsible – the driver (possibly the only “passenger” at the time of the incident), or the car manufacturer?
It is safe to say that autonomous driving is anticipated by costumers and car manufacturers alike (read more about what consumers want in Simon Parry’s blog, “Self-driving cars: what do consumers think (and want)”. However, given the current uncertainties, how can the transition to fully autonomous driving be best managed?
Some manufacturers, such as Tesla, believe in directly implementing fully autonomous driving (the technical aspects are discussed in Ben Boyd’s blog, “How self-driving cars are putting the brakes on vehicle collisions”). They believe fully autonomous driving is a unique selling proposition which emphasises their technical leadership and attracts new costumers eagerly waiting for this functionality. In fact, Tesla claims that their current models could be self-driving cars if they were allowed to switch on this functionality, i.e. the legal framework were to allow self-driving cars.
Other manufacturers favour a different approach. They consider firstly developing and implementing semi-autonomous driving (SAE level 2 or 3) and using this technology as a starting point for fully autonomous driving. There are good reasons for this approach. Firstly, the implementation of fully autonomous driving is of no use if the regulatory environment does not allow for fully autonomous driving or the regulatory provisions differ from country to country. Secondly, it might generate more revenue if the SAE level 2 or 3 driving can be sold in the immediate future, instead of waiting for fully autonomous driving down the road. This approach is especially relevant for traditional car manufacturers who may be concerned about the financial capacity to simultaneously transition from combustion engines to the electric world.
In order to alleviate the costs and risks involved in this process, traditional car manufacturers have proven to be open to new and sometimes surprising partnerships. For example, longstanding rivals Daimler and BMW have launched a development cooperation on automated driving. As semi and fully autonomous driving heavily relies on software applications, some car manufacturers (e.g. Volvo) collaborate with software companies such as Google’s autonomous driving specialist Waymo.
Given that autonomous driving is a relatively new development, long-term experience concerning the reliability of self-driving systems is lacking at the moment. For example, it is unclear whether sensors/cameras for analysing the environment of the car will deteriorate over time and how this will affect the robustness of the algorithm for autonomous driving. What is more, it remains to be seen how self-driving systems will be certified in terms of road safety. Car manufacturers are understandably reluctant to disclose the algorithms of their self-driving systems. However, this means the software is a black box for any road test of the self-driving car. A balance between legislators concerned with road safety on the one hand and the legitimate interest of the car manufacturers regarding their intellectual property on the other hand needs to be found.
In any case, the development of the SAE level 4 or 5 technology is not separate from the development of semi-autonomous driving. This also applies to the development of the regulatory environment needed to cater to the different interests of customers and car manufacturers. Exciting times are definitively ahead and we, at Mewburn Ellis, will continue to play a part in this developing technology which will affect us all.
