Autonomous driving is currently the big dream in the automotive world. And sometimes, just after getting up in the morning, I find myself thinking about getting into a self-driving car after eating my breakfast and letting it take me to the office. Thankfully, my morning coffee quickly brings me back down to earth.
Away from dramatic, attention-grabbing headlines and campus shuttles barreling along at 15 km/h, it can be said that the race for the first Level 5 car is at full tilt. But as the German saying goes, “life the development of the autonomous car is a construction site.”
The complexity of autonomous driving is difficult to imagine. We are expecting an admittedly elaborated (originally mechanical) construction made of sheet metal and rubber to develop its own thoughts – at least to the extent that it will be able to bring us safely to our destination in a busy, or even hazardous, environment. It reminds me a little of Rabbi Loew’s Golem – originally just a lump of clay, shifting through the streets of Prague as a silent protector. Some people find navigating Germany’s city center streets difficult – and now we are expecting a car to be able to do it by itself.
And there’s something else: So that autonomous vehicles are at least as good as humans, they need to be able to think outside their own box. They have to “see” more than just the data in their on-board sensor technology. They have to converse with smart traffic infrastructures, and ideally with other road users as well. So autonomous driving is about more than just an autonomous car. The environment has to be smart too, and that means collecting data.
For the moment, it has nothing to do with 5G and edge computing, even though they are, of course, both important components that generate additional intelligence, particularly within the scope of cellular Car2X communication. Embedded systems and 4G can solve many of the challenges of autonomous driving even today. The real problem lies not in the technology, but behind the scenes. As we know, the development of the autonomous car is a construction site.
At the moment – how could it be any different? – there are many different workers at play, and therefore many different technological aspects: excavators here, cranes there, cement mixers here, casings there. And like a large construction site, autonomous driving requires central coordination. Today, this should take place through an ecosystem, or its technical mapping – a joint platform. The key requirement for the success of autonomous driving is the transfer of domain knowledge from previous system integrations and big data scenarios into operating platforms for autonomous driving.
Alongside over-the-air updates, such an integrated operating platform can make an important contribution to the monitoring of all components of the ecosystem, or for automotive-specific security solutions. On top of that, this operating platform requires automobile manufacturers to have an open mind when it comes to the engineering process. For example, the digital twins of vehicles from a development and operating perspective can be combined with the digital twins of their environment so that the operating results can feed directly into the design process.
The fact that such operating platforms also need real components for authentic tests is a given. Such an endurance test is currently taking place on the street Straße des 17. Juni in Berlin. In the autonomous test field, autonomous cars travel in city traffic through a digitalized environment.
