Mankind can’t resist a challenge. As with the first landing on the Moon, it will be the same for driverless cars – we will get there eventually.
Where scientists, engineers and explorers conquer, however, big business and consumerism will follow and these days that also means internet reviews and YouTube videos. The court of public opinion and confidence has to be continuously satisfied – amply illustrated by one recent review of a driverless experience that said “…it seems that the future of driving is equal parts glee and terror”.
In the UK and elsewhere, truly driverless cars are still being officially tested and will not be available to the public for many years. To bridge the gap, motor manufacturers are developing and releasing advanced driver assistance systems (ADAS) that allow fully driverless capability in given situations. The latest example in the USA is Tesla’s Autopilot system, itself still at the beta-testing stage, which like other technologies at this stage, has already been the subject of internet video, media coverage and opinion.
Any on-line video of ADAS is impressive but it also brings a reality check that balances the sales pitch. Alongside advertising videos extolling the virtues is footage of so-called “fails” and those already released by consumers highlight the potential for conflict between the driver and the manufacturer in the event of an accident: it was the fault of the car says one; no, you weren’t using it properly says the other.
In the case of Autopilot, for example, Tesla expressly directs drivers to keep hands on the wheel at all times, even when the vehicle is driving itself. Tesla is clear that responsibility will at all times rest with the driver.
The need for such caution is made clear by two recent videos of ADAS in action. In the first, the driver is not completely at ease; after safely driving “hands-free” on a dual carriageway for some minutes, and with hands now hovering over the wheel, the vehicle takes a slip lane. It does not follow the left-hand curve as expected and heads for the curb to its right, causing the driver to intervene. In the second, a car first veers towards the opposite (empty) carriageway and is corrected; moments later it again veers towards the opposite carriageway but now also an oncoming car and an incident is immediately averted by the driver.
The point here is that even sophisticated ADAS will have limitations which drivers will need to understand but may not, and manufacturers may find they need to do more to avoid drivers having a false sense of security. Questions that could in future be asked of any vehicle manufacturer are:
- were these road conditions within the design parameters and capability of the ADAS?
- if not, had the driver been given and ignored instructions on proper use of the system?
- if the conditions were not within the capability of the ADAS, should they have been?
- can a manufacturer reasonably say that road type A is safe for ADAS but road type B is not safe when – potentially – a driver won’t immediately or quickly enough be able to identify which road type they are on?
Watching real people use ADAS also illustrates, from the differing responses, how important driver training is going to be in future. Reactions varied but two stand out – in one case the driver was so intent on observation of the driver interface screen that awareness of the general road conditions clearly became secondary; in another the driver was so confident he climbed into the backseat (despite the screams of his now hysterical passenger….).
A final observation is that the increasing use of dash-cams or other on-board cameras is undoubtedly contributing to the availability of real-life feedback in this area and should be encouraged. What began for many people as a personal counter-fraud measure may, in fact, become an integral part of the connected and driverless car experience, to the benefit of all.
Written by Nick Rogers, partner and head of the BLM motor group