Putting driverless cars on the map: An interview with an AD mastermind
He is one of the most powerful men in the automated driving business: Russ Shields. In an exclusive interview with 2025AD, the U.S. entrepreneur explains why navigation databases will save lives – and why he is not a fan of Tesla.
Russ Shields is not a man who seeks publicity. Interviews with him are rare. That is why there are even automotive experts who have not heard of the U.S. businessman. But don’t be fooled by his media absence: when it comes to automated mobility, Shields is a big presence. His company Navteq evolved into HERE – a key player in navigational maps, now owned by BMW, Audi and Mercedes. Today, his company Ygomi is developing complex solutions for road network databases – a prerequisite for autonomous driving.
2025AD: Mr. Shields, you are known as a gambler. During your college time you were a successful poker player. If you had to bet on the year we will see fully autonomous series vehicles on our roads, which year would it be?
Russ Shields: Cars with SAE level 4 automation, which is highly automated driving, will probably be on the market in 2020 or 2021 – on limited expressways. By 2025, there will be more and more roads covered, possibly even country or mountain roads. But a true door-to-door autonomous car on all public roads – SAE level 5 automation – is many, many years away. We are probably talking about the 2030s.
2025AD: What you say is in line with the position of the big German OEMs. Some American companies like Google sometimes give the impression that they will be able to pull off a level 5 vehicle that can drive on any roads in the next couple of years. Are you more skeptical about that?
Russ Shields: Google has gotten under pressure recently. They are now focusing on very constrained areas, not all roads. And even these won’t become a reality within the next couple of years. Similarly, Ford is eyeing 2021 for a shuttle vehicle in a closed area – not something that is for widespread sale. And I don’t hear anything different from Ford, GM or the Japanese companies like Toyota or Nissan from what the German manufacturers are expecting over the next ten years. Uber might be pressing for a quicker approach but I’d be surprised if they could get an SAE level 5 vehicle on the road in the near future. They just don’t know how hard it is.
2025AD: You are someone who likes to take a clear stance. In 2015, you dismissed the Google car as “a joke”. What led to your opinion?
Russ Shields: I still think so today. But they have moved their objectives. Now they are saying they’re not really building a car. Instead, they plan to license technology to vehicle manufacturers because they do have some very good technology. This makes their work more practical.
2025AD: One of the key prerequisites for automated cars is digital maps. You are one of the pioneers of digital mapping technology. Under your leadership Navteq became the world’s most comprehensive navigable database. At what point did you realize how important this technology was for automated driving?
Russ Shields: We did early pilots in the late 1990s on how electronic maps could improve safety systems like adaptive cruise control with lane handling. At that point, there was no serious thinking about highly automated driving. It’s only really been in the last few years that the potential became obvious.
2025AD: Navteq has had quite an interesting career. It was sold to Nokia, then renamed HERE which was then bought by Audi, Mercedes and BMW. How do you assess the future prospects of the platform?
Russ Shields: The crucial question is: Will they be able to create the software necessary for the machine part of future vehicle generations? They have good experience and a good market share. But the technology requirements are evolving. Mapping technology for humans – like navigation and telematics – is very different from the machine-oriented technology which we need for SAE level 5 automated driving.
2025AD: What is the difference?
Russ Shields: The so-called road network databases for autonomous cars require a much higher level of accuracy and reliability. Mistakes in navigation systems happen all the time. But for the machine-to-machine uses of the database, mistakes in navigation systems can end up killing people. Avoiding that is the challenge that HERE and other map companies face. How will they keep their very successful human-oriented business and create the complex, highly reliable software for the machine-oriented part of the business at the same time?
2025AD: There are several other players besides HERE who work on road network databases. Will there at some point be a need for those mapping giants to create common standards?
Russ Shields: To create standards for a complicated technology before it gets into the market is pretty hard. We are a long way from even having standards for chip sets. So it’s unlikely that we will have any kind of successful software standards even long after the systems are on the market and reliable.
2025AD: Since you are an expert for Asian markets, do you see any specific challenges for digital mapping in this region?
Russ Shields: The database needs to accurately reflect what’s on the road. Every area of the world has different conditions because they have different land uses, different population structures. So the software has to be tested and carefully adapted to each market. Additionally, there is no assurance that there won’t be interference by governments wanting to protect national businesses. That’s a risk that always exists.
2025AD: Your company Ygomi delivers innovative software and technology-based service for automotive manufacturers. Where does your focus lie when it comes to automated driving?
Russ Shields: Our major target is to build a road network database for automated driving. We have been in the software business for 47 years and there has always been technological evolution. But today we can provide software that allows for things that were not at all possible and economically practical in the past.
2025AD: What new possibilities arise from this technological progress?
Russ Shields: For building a road network database, processors are now becoming fast enough to support highly automated driving in high-end cars. Eventually we will even get into mid-price and low-end cars. I expect that over the next 10 to 15 years, every new car built will include a road network database to support the safety systems.
2025AD: It seems like the media often focuses on a competition between German OEMs and Silicon Valley players. Do you think that the media underestimates the competitiveness of other regions, Asia in particular?
Russ Shields: I see a lot of potential in many Chinese companies with huge financial capabilities. They may evolve to be a successful vehicle manufacturer on the world stage – similar to the way that Japanese OEMs have become world-class competitors. Automation and electric powertrains will play a huge role in that regard. 20 years from now, I wouldn’t be surprised to see half of the world’s existing ten largest manufacturers having gone out of business with new ones having taken over their roles.
2025AD: With the recent incidents involving Tesla’s Autopilot, one could almost get the impression that autonomous driving is at a crucial point where trust in the technology is being tested…
Russ Shields: Even before Tesla released the Autopilot, many conservative voices said that automated driving is going to be very difficult to deploy – and that there will be cases where inadequate systems will be brought into the market. Then the release of the Autopilot raised a lot of questions. Many experts were saying that it was too early and not ready to be put into the market.
2025AD: What led to their criticism?
Russ Shields: Here’s one example: Tesla’s manual says that automated braking will only work if the object you are braking for is moving – and that you will hit a fixed object. That’s a level of incomplete technology that I would not expect most OEMs to allow in their products. I am familiar with at least five accidents that have happened with the Tesla Autopilot that have been publicly registered. Only two of those were fatal. But for the number of cars they have on the road, that’s a pretty high accident rate. Hopefully most car companies will be more careful in what they do then Tesla has been.
Will Chinese OEMs play a big role in automated driving? Will it be the 2030s before Level 5 automation hits our streets? What do you think of Russ Shields' assessments? Share your thoughts in the comment section!