Ford CEO Mark Fields announced yesterday that the automaker will produce a fully self-driving vehicle by 2021 that can be used for automated ride-sharing services. Fields made clear that the vehicle will have Level 4 autonomy “without a steering wheel and pedals” and will be “mass-produced.”
Meeting these goals is largely the responsibility of Randy Visintainer, director of Autonomous Vehicles at Ford. I spoke with Visintainer after the announcement to find out not only how he and his team plan to meet the five-year deadline to bring a fully autonomous car to market and what the production entails, but possible future partners and where the first Ford self-driving car will launch.
Newcomb: Tell me what yesterday’s announcement means for Ford and directly for your job?
Visintainer: I came back to lead the vehicle program in technology development a year and a half or so ago. It had been a research project and we were transitioning from the research phase to what we call the advanced phase. The research phase is where things generally begin and we’re entering a very high level of “can we do this.” We formulate a hypothesis and test the hypothesis. If we answer the question affirmative or the hypothesis looks valid, then we go to the next phase, which is the advanced phase. So that gets into: “We know it’s feasible if we do it. Now how do we actually do it with suppliers? What does the technology need to look like? How we’re going to develop and validate the technology? What kind of vehicle platform would this be aligned with?” And so on. The next phase following the advanced phase is integrating it into a vehicle program. And the final phase is manufacturing that product.
Newcomb: What stage are you at right now?
Visintainer: We’ve transitioned into the advanced stage, but because the technology development and the vehicle development are so intertwined, we’ve kind of blurred [those lines]. Instead of being sequential, we’re parallel. We’ve got a program team up and running and we’re looking at the requirements of the vehicle platform as we’re evolving the technology. In addition to developing the technology and laying out a vehicle program, we’re also looking at the unique manufacturing requirement for this vehicle. How we’re going to mass produce this vehicle in an assembly plant. That’s an important work stream as well.
Newcomb: Is this a clean-sheet approach, in that Ford is creating an entirely new vehicle or is it using an existing platform?
Visintainer: We’re in the pre-program phase, which is starting to define the specifics of the program. That’s why it makes sense to run [the technology development and the vehicle development] in parallel. Because how we define the requirements of the vehicle are so tied to the technology in this case. What we are looking at is leveraging [Ford’s eight vehicle] platforms and which ones make the most sense for the size of the vehicle. We’ll probably try to leverage as much as we can of the existing and upcoming platforms that will be available to us. But there’s going to be substantial differentiation between this vehicle and the others that are on that platform.
I can give you an example of one of the fundamental differences. In a vehicle with controls and a human driver, the driver is the backup for some of the functionality of the vehicle. For example, the electronic steering. If the electronic part of the steering fails, there’s still a mechanical linkage and the driver can still manually steer the vehicle. When you take the driver out of the loop and there are no controls in the vehicle, then you have to replace that with a redundant steering system. And so the steering system for an autonomous vehicle is different than any other system that we’ve got in production today. And the brakes are a similar example.
Newcomb: So the vehicle being developed without any manual controls?
Visintainer: Right. It’s a Level 4 vehicle for ride-sharing service, so the intent is no manual control. We have to provide those backups because we don’t have a human to rely on.
Newcomb: Yesterday, Ford executive vice president of product development Raj Nair touched on the regulatory environment surrounding autonomous vehicles, which is a bit behind compared to the technology. Any comments?
Visintainer: It’s a very dynamic situation and we’ve seen there’s been a lot of activity. We joined the Self-Driving Coalition for Safer Streets. We see that autonomy is going to help people travel safer and more efficient and facilitate mobility for those unable to drive. So we’re working with some cross-functional partners like Google and Uber to advocate for policies to support the development of autonomous vehicles.
Newcomb: Is the five-year timeline a big challenge?
Visintainer: It is a challenge, but we believe it’s achievable or we wouldn’t have gone public. Both Raj and Mark made reference to that we’ve been working on a research level in autonomy for over 10 years. And we’ve seen the technology maturing: advances in the machine learning and the compute platform, the speed and density of the silicon, the sensor capabilities and costs coming down. As we’ve been moving our research forward and taking advantage of these advances, we felt confident we were ready to move into the advance phase of the project last year And we’ve been building confidence around this timeline over the last year.
Newcomb: As we see the project unfold, will other partners be announced?
Visintainer: We are talking with other partners. And as we continue to move forward with the programming and we follow the process, they will be announcements about who we’re going to work with in some of the new technology.
Newcomb: How large is the team working on this?
Visintainer: The great thing about being at Ford is we leverage core engineers. So we have chassis and body and electrical people working on this that are part of core organizations at our vehicle operations. So we’ve got well over a 100 people engaged in this, in addition to the core team of roboticists, artificial intelligence machine learning people and specialists that are coming up with the redundancies in the systems. So we’ve got a very large team between the core team and the focused AV team.
Newcomb: How much of the work is going to take place in Silicon Valley and how much will take place in Dearborn?
Visintainer: Most of our work right now is done in Dearborn. We do have a team in Palo Alto who are all part of one global team. We also have a team that is located in Aichen, Germany. So we have a global work load and based on where the skills are, we distribute that work load across the regions to Germany, Dearborn and Palo Alto. And we’re growing the team in all three regions.
Newcomb: It sounds like Ford is looking at this as a global platform as opposed to a certain segment like North America or Europe?
Visintainer: That’s correct and that’s part of our platform strategy – that all of our platforms have the potential to be global platforms. So we are looking to make sure we could leverage this platform globally, but the initial focus is North America. That is our current focus and all of our development work is focused primarily here in North America.
Originally published to Forbes.com