Two years since Daimler Trucks first tested its self-driving truck — the Inspiration concept — competitors have jumped into the arena.
Uber Technologies completed an autonomous 120-mile shipment last year. Alphabet’s Waymo recently began testing as well. Tesla Inc. and Nikola Motor have both committed to developing electric trucks with self-driving capability. Volvo Trucks is deep into testing automated technology and sent a self-driving truck to work in an ore mine.
But Daimler is taking a measured approach to the technology, Derek Rotz, director of advanced engineering for Daimler Trucks North America, told Trucks.com in an interview.
For example, when Daimler’s Freightliner division unveiled its 2018 Cascadia truck last year, the vehicle came with some advanced technology functions such as automatic braking and lane departure warning but is nowhere near self-driving.
The Inspiration test, Rotz said, taught Daimler that the road to fully autonomous driving, is still fraught with obstacles.
Here is what Rotz has to say about self-driving trucks and Daimler’s strategy.
Will your department become more important as rivals develop autonomous technology in trucking and shipping?
Absolutely. It was 2015 when the Inspiration truck came out, and that was a big milestone for the industry, having the world’s first licensed autonomous truck for testing in Nevada. Since then it’s been a lot of very public activity coming out of not only the automotive and trucking industry but also the tech industry so it’s a very interesting time.
Our role is to look with a further time horizon and investigate different innovations that can help improve the vehicle in terms of fuel economy, safety and connectivity.
What is the state of the Inspiration project?
The goal of the Inspiration truck we demonstrated in 2015 was to basically showcase the technology that could get us to autonomous driving. What we’ve learned is it’s still a long road ahead. The vehicle has to be able to bring itself to a safe state if the driver doesn’t re-engage, or things break. A rock chip will hit a sensor and damage it – you name it. There’s a number of things that can happen. These systems need much more redundancy, much more development to make that step to autonomous driving.
Does the technology need more work?
The technology is there to basically showcase it. But to go from, what we call a minimum viable product, to a commercially viable product means making sure that it works 100 percent of the time — not just once. To do the demonstration is certainly challenging in its own right and it’s a big feat, but it’s the first few steps on a long hike to the mountaintop.
What technologies have made the most progress in the couple years since the project began?
What’s improved is actually something called Sensor Fusion. It’s taking the best aspects of the radar system, taking the best aspects of the camera system and merging those together. What we’re seeing is fusing the sensors from those different sensor sources gives us a better picture of the information that’s out there. That allows you to do things such as pedestrian detection.
How much information gets shared between Daimler’s international operations?
That’s the beauty of our global setup with Advanced Engineering. We do work closely with our counterparts in Stuttgart, Germany as well as in Kawasaki, Japan on these topics. Literally the engineers are working around the clock with each other on these topics, it’s a big enough topic that it doesn’t make sense to develop it three times in three different continents. It’s a global development.
Is Advanced Engineering involved with the company’s Mercedes-Benz passenger car division?
Daimler Trucks North America is very fortunate to be able to leverage the development that is being done for passenger cars because that is certainly very relevant to what we’re doing in trucks. Being that a lot of it is dealing with intelligence and software, those are things we can benefit from. We monitor that very closely. We see that as actually a distinct advantage that we have is to be able to leverage that Daimler warehouse research, so to say.
What is the current state of autonomous technology advancement?
Autonomous does hold the promise of improving safety, improving fuel economy, improving driver comfort and productivity. In the late 90s we put anti-lock brakes in trucking and that was our first active safety system. Fast forward a number of years before we put active brake assist on and the camera systems for lane departure warning and that’s where we’re at today.
It’s a challenge for me to see how that would pick up a very quick steam in terms of development. Between the spot where we’re at today and the spot where autonomous driving is, that hill is actually a fair bit steeper in terms of what we need to do.
Because the technology hasn’t progressed as quickly as you thought?
You start discovering something we call “edge cases.” These are driving situations which are anomalies and sort of unique but maybe wasn’t thoroughly thought through beforehand. Like detecting road debris for example. Poor lane markings, that’s a big one actually. What happens if a steer tire blows out? Humans are very good at intuitively assessing a situation and responding. Computers need to have all that lined up for them, all those edge cases. I get a little winded even thinking about it.
Where does platooning fit in to the hierarchy of technologies?
Platooning does leverage those technologies such as cameras and radar systems. It does add a new component on what is called vehicle-to-vehicle communications. The V2V signal is broadcasting the braking status 10 times a second, so that following vehicle can respond within maybe 200 milliseconds to the vehicle in the front. And that coordinated braking actually is a key component of platooning and also a feature that I believe can bring some safety benefits.
As with autonomous driving, in platooning there’s a patchwork of rules state by state that aren’t uniform right now. We follow that and look forward to a day where the rules would be in place on a state by state basis which are uniform that would allow platooning to happen.
What are the current technology development goals?
We want to improve active safety. We want to decrease the number of crashes. We want to make sure that the vehicles on the road are going to be the safest out there and that every year we put new vehicles on that they’ll be safer than the ones before. That’s really our focus.
I would imagine the Cascadia is a big step in that direction.
Absolutely. The Detroit Assurance Package which is our safety suite that’s integrated within our powertrain does a great job with adaptive cruise control, the active brake assist and the lane departure warning system. Essentially, we’re going to take those systems and just keep building on it and make them even better tomorrow.