Leave the “we promised you autonomous vehicles” World behind. Get validated in the real world.

Lately I was in Berlin, twice. First as speaker and panelist at the "Autonomous Vehicles Europe", an international conference with 50+ participants with a broad background in “autonomous”, from both inside and outside European Union, most notably Norway.

And no, they did not bring the icebear, it was already here, part of the “Bear-lin” branding:

Second, a few weeks later, I got invited by three federal ministries, to the "Fachkonferenz Autonomes Fahren” (Autonomous Driving Conference), where representatives from current "autonomous" research projects in industry, research and education, but also from regulatory bodies, administration and regional mobility providers – close to 300 participants - came together from all over Germany.

One of the main topics at both conferences was the trend towards the use of “autonomous” vehicles in Mobility-as-a-Service (MaaS) schemes. This is less about additional mobility offerings in the city and more about connecting rural areas and suburbs that cannot be served economically with existing technologies.

General agreement was that in 2023, we have the technology in our hands to drive on good roads, with highly automated vehicles, from A to B. Self-driving within strictly enforced ODD (operational design domain) is feasible. Still work to do is on certification and validation of algorithms and autonomous driving in general.

So, discussion point for many projects in Germany was how to actually accelerate the current approval procedures, in line with the recently introduced “Genehmigungsbeschleunigungsgesetz” (nice german word for a law to accelerate approvals). Our current legislation favors manufacturers with large quantities, only.

What we found is needed most? "Mehr MUT" - More courage!

In general, the rampant risk aversion got criticized, and Andreas Reschke from pony.ai gave a number of reasons why “autonomous” driving in Germany currently is "not worthwhile for an operator of robotaxis, or logistics solutions. We are a commercial company".

With our extensive rules and regulations, we do have a big advantage when the TÜV (technical regulations body) comes knocking on the door of foreign companies. A few of the larger “autonomous” players quit the market during the course of 2023 for a number of reasons, not least because of validation requirements.

Disclaimer: my sponsor to bring me to these events, ViGEM GmbH (www.vigem.de), designs and builds robust and reliable data recording equipment for use in development and validation of ADAS.

Next, the conferences in Berlin also touched one of my focus fields in 2023 .. the Software Defined Jungle. In 2023 everyone and their oncle put SDV on their banners to catch the SDV train but only a few will advance into the elect field of serious contributors.

What is crystallizing out of the many discussions I had with experts from the software- and automotive-industry this year: there are too many parties [hence my “software defined jungle”] offering a helping hand. OEM acceptance towards a common, open way forward is still missing - "open SDV" needs to get a few OEM onboard to get the stones rolling into the same direction!

Until recently, the in-sourcing activities by a number of large OEMs to build their own software factories, much like they used to have their own machine shops to build their own combustion engines, did not achieve the goal of software ownership. Software engineering does not work that way.

Already a good part of their software was taken from open source projects, and further intitiatives towards isolation of their software’s unique selling points within their software framework made many minds realize that there is not much left without giving customers a say in how they want to use the user-relevant software configuration in their vehicles.

Rather, a common base-software and mostly common middlewares will help the industry to come up with standardized application interfaces (API) that can be used by third parties apps without much hassle from an OEM who runs a particular vehicle platform.

But, “this will take money away from the OEMs!” they cry. Who thought to be able to force historic business models onto customers in the 21st century? The traditional model of individual car ownership is waning in the face of proliferating car-sharing and mobility-as-a-service offerings.

And, when I pay for heated seats, I want to be able to use them, regardless of any actual car brand under the back of my pants.