Never miss a story from Breakfast Bytes. Subscribe for in-depth analysis and articles.
You hear a lot of talk about autonomous cars, but I've also heard many times that autonomous trucks may well be first. If it costs a fixed amount to make a vehicle autonomous, then it is much easier to absorb that cost into a $150,000 truck than a $20,000 car.
Anyway, at Ludwigsburg, Dr Rainer Müller-Finkeldei (doesn't that sound like the sort of name Mike Myers would come up with for a German spy in an Austin Powers movie?) of Daimler talked about The Intelligent Truck—Safe, Autonomous & Connected.
Approximately 90% of all truck-related accidents are caused by human error. Of all accidents involving trucks (these numbers are for Germany, but nobody seems to think other countries would be significantly different):
Many of these can either be avoided or mitigated by technologies that we already have: blind-spot assist, active brake assist, and lane-departure warning. Many high-end cars have this technology, sometimes called AEB (for automatic emergency brake) and lane-departure. Blind-spot assist is not really necessary in a car since the blind spot is tiny compared to that for a truck.
However, Rainer pointed out the challenge to going to the next level. This type of automation is catching things that people do wrong, catching more and more of the errors that humans make. But humans do much more right than wrong when driving and future automation is about things that people do right. The big challenge is not to make errors when humans would make none.
The next stage, which is coming soon, is autonomous driving (still talking about trucks, of course) on highways. This is not a truck driving itself with no driver. The driver remains in charge and is still responsible.
Things that need to be handled between the autonomous systems and the driver are:
This "Highway Pilot" requires attention monitoring (of the driver, no sleeping) and dedicated handover timings. It also would be desirable to change the law to extend the legally allowed activities while driving, and, due to reduced stress, extend the legal "driving" hours. The simple case of driving on a freeway in normal traffic or in a traffic jam can be handled by the systems. But construction zones and exit ramps require the driver to take over, at least in the coming generation of autonomy.
The next stage is to add V2V (vehicle-to-vehicle) connectivity, allowing platooning. The idea is to make a sort of train out of multiple trucks. The driver and systems in the front truck are the "engine" of the train. The other trucks largely follow the one in front. However, even the drivers in the trailing trucks see out the front of the whole platoon via camera.
But stuff can happen in a platoon that doesn't happen in a train:
The experimental results are enhanced safety for all road users and fuel saving of ~7% (3% for leading vehicle and 10% for trailing vehicles). The trucks can be more closely spaced, which makes better use of the roads. Even though every truck still has a driver, there is a lot less strain on the drivers.
The big question is how to enter cities in trucks. There is an expectation that some (all?) cities will eventually ban diesel trucks, and so the solution is going to involve what Daimler calls e-mobility, some form of EV. But trucks are a lot bigger than cars so the tradeoffs between battery capacity, range, payload, and cost are not the same.
The above picture is the first Mercedes-Benz urban eTruck. It has not been custom designed, in the sense that they used an existing truck design, took out the diesel traction and added all the stuff you see. But the cab still has relatively restricted vision since it is designed to tilt forward for access to the now non-existent engine,.
It has 212 kWh of power, giving it a 200km range and 12.8 ton payload (which is still too low).
Trucks are used just 35% of the time. Standing or parked is nearly 50% of the time. Loading/unloading 5.5%. Waiting 9%. Administration 1.5%. Also, every journey by an empty truck is adding to congestion with no net-benefit (apart from the obvious one of repositioning an empty truck).
Mercedes-Benz (and other truck manufacturers) are adding fleet management systems in the cloud, and using real-time data on truck positioning, to improve utilization and get that 35% of "useful" time up. Data is the new oil, as several people said during the Kongress.
Of course, all this truck technology is benefitting a lot from what is going on with passenger cars. Not just technology, but passenger car volumes will drive down the cost of batteries and other technology such as lidar. However, for trucks, Rainer believes there will be no competition from ICE since cities will ban non-electric trucks.
My prediction is that we will see a hybrid pilot model. This has nothing to do with aircraft pilots, but with ship pilots. When a big container ship arrives outside the Golden Gate, a pilot goes out to the ship and takes over. He or she knows San Francisco Bay intimately and guides the ship to its destination. Long-distance trucks are similar in that they go for hundreds if not thousands of miles on the interstates, and then need local knowledge to get from, say, the Bay Bridge to a Safeway loading dock. It is much easier to automate the interstate driving, than it is driving a large truck around San Francisco and backing it into a loading dock.
For large tractor-trailer trucks, I think we will do the long-distance driving with diesel engines and the technologies discussed in this post. Then, at the outskirts of the city, the tractor will be switched over to an electric tractor. But the driver will be switched over too, to a "pilot" who knows the city, takes the trailer to its destination. Obviously, in the reverse direction, the pilot and electric tractor takes the vehicle to the edge of the city and then the diesel tractor takes it out onto the highway to drive across the country.
Trivia fact of the day #1: Janis Joplin recorded this on October 1, 1970. She died on the 4th.
Trivia fact of the day #2: The song title on the album (Pearl) has no hyphen between Mercedes and Benz although, of course, the vehicle company does—so the title of this post is correct.
Sign up for the weekly Breakfast Bytes email: