Q&A: Dr Dirk Hoheisel

Bosch's member of the board of management explains the advances in connectivity and autonomous vehicle development

The trend is towards more automated functionality, and ultimately fully autonomous vehicles, but where is the push coming from?
It’s a very simple answer – safety in the car. We see that the numbers of fatalities on the roads are significantly reduced by electronic systems, but we also see in developed markets a stagnation of accident figures, or them slowly increasing again.
Our vision is zero fatalities, and to reach this we have to bring automation inside the car to make it more reliable. And in the future we believe that autonomous cars have the possibility to reduce significantly, by magnitudes, the number of fatalities.
And to use your vehicle as the third living space is also the motivation. In traffic jams you use your smartphone, but it’s better you do that in an autonomous car.

Bosch has developed its highway pilot technology. How big a leap is it from this to a fully autonomous vehicle?
The highway pilot handles traffic situations that aren’t too complex – on highways the traffic typically is in one direction and there aren’t so many disturbances and unforeseen issues. I think urban situations are the most complex. We could have very low-speed vehicles, but I don’t think people would accept this solution because then you would have non-autonomous cars overtaking these creeping vehicles. But although I think it is an enormous step, it isn’t only technology that poses challenges: we also have legal issues that have to be solved. At the moment, especially in Europe with the Vienna convention, we can’t even park autonomously, and we believe the next big thing is autonomous valet parking. So first the politicians have to change the rules, which they are working on, which is a good message, but at the moment we aren’t able to do even this.

How advanced will sensor technologies need to be to meet the requirements of autonomous vehicles?
At the moment we have one-megapixel cameras in series production. We believe that we’ll need more information, for example to identify road signs or traffic lights, so we’ll need a higher resolution to receive stable information. So the next step in our roadmap is to also have two- and four-megapixel cameras.

Any increase in the amount of sensor data will presumably mean a boost in processor performance too?
We have multi-core chips and we will introduce many more cores. It’s the same story as in consumer electronics.

More vehicles now benefit from head-up displays, and the screen technology is increasing in size, which appears to be leading us towards augmented reality – is that a definite trajectory?
In general, yes, we are also working on this technology and it’s a clear trend. It’s really a challenge though because of the packaging of these bigger screens and the optics behind them.

Increasing the amount of information given to the driver throws up many challenges too. How do you meet requirements in terms of usability and distraction?
We have more information that we want to share with the driver, especially with automated driving functionality, because the driver must be sure that the vehicle has everything under control. This is also the reason why we developed head-up displays, because this is a natural location to display information. But I think we also have to look to consumer electronics technologies. For example, gaze control – it’s very fast. We have to work on the technologies that distract the driver less.
This comes from consumer electronics, the games industry for example, and we adapt these and bring them into our cars relatively quickly.
If you look back in history, with cassette players or CD players, it took time for these technologies to move into the car. But today there is really no time difference between introduction in consumer electronics applications and in the car. It’s shrunk significantly.