Nikkei: Japan carmakers now rely on emerging markets for 62% of profit
Carbon nanotube–graphene complexes show promise as low-cast, high-activity electrocatalysts for fuel cells or metal-air batteries

SARTRE road train premieres on public roads; focus now shifts to fuel consumption

SARTRE public road test May 2012
SARTRE public road test. Click to enlarge.

Vehicle platoon tests in the SARTRE (Safe Road Trains for the Environment) project—a joint venture between Ricardo UK Ltd, Applus+ Idiada, Tecnalia Research & Innovation, Institut für Kraftfahrzeuge Aachen (IKA), SP Technical Research Institute, Volvo Technology and Volvo Car Corporation—are making progress. (Earlier post.)

For the first time a road train comprising a Volvo XC60, a Volvo V60 and a Volvo S60 plus one truck automatically driving in convoy behind a lead vehicle has operated on a public motorway among other road users.

While there remain many challenges to full scale implementation, the SARTRE project has demonstrated a very practical approach to the implementation of safe road train technology that is capable of delivering an improved driving experience, better road space utilization and reduced carbon dioxide emissions.

Once the fuel consumption measurements are completed we will be drawing on the learning we have gained developing the platoon system and understanding the various human factors, to assess the likely roadmap and mechanisms for platoons and platoon technology to be operational on public highways—at which point we believe there will be a really positive impact on highway utilization.

—SARTRE project director, Tom Robinson of Ricardo

A road train consists of a lead vehicle driven by a professional driver followed by a number of vehicles. Building on already existing safety systems—including features such as cameras, radar and laser sensors—the vehicles monitor the lead vehicle and also other vehicles in their immediate vicinity. By adding in wireless communication, the vehicles in the platoon “mimic” the lead vehicle using Ricardo autonomous control—accelerating, braking and turning in exactly the same way as the leader.

The project aims to deliver improved comfort for drivers, who can now spend their time doing other things while driving. The project also aims to improve traffic safety, reduce environmental impact and cut the risk of traffic tailbacks.

The vehicles drove at 85 km/h (53 mph). The gap between each vehicle was only six meters; during trials on the test circuit, the researchers tried out gaps from five to fifteen meters.

The three-year SARTRE project has been under way since 2009. All told, the vehicles in the project have covered about 10,000 kilometers (6,214 miles). After the test on the public roads in Spain, the project is now entering a new phase with the focus on analysis of fuel consumption. The estimated fuel consumption saving for high speed highway operation of road trains is in the region of 20% depending on vehicle spacing and geometry.

We covered 200 kilometeres in one day and the test turned out well. We’re really delighted. Driving among other road-users is a great milestone in our project. It was truly thrilling.

We’ve learnt a whole lot during this period. People think that autonomous driving is science fiction, but the fact is that the technology is already here. From the purely conceptual viewpoint, it works fine and road train will be around in one form or another in the future.

We’ve focused really hard on changing as little as possible in existing systems. Everything should function without any infrastructure changes to the roads or expensive additional components in the cars. Apart from the software developed as part of the project, it is really only the wireless network installed between the cars that set them apart from other cars available in showrooms today.

—Linda Wahlström, project manager for the SARTRE project at Volvo Car Corporation

Part-funded by the European Commission under the Framework 7 programme, SARTRE is led by Ricardo UK Ltd. SARTRE aims to encourage a step change in personal transport usage through the development of safe environmental road trains (platoons). The objectives of SARTRE may be summarized as:

  1. To define a set of acceptable platooning strategies that will allow road trains to operate on public highways without changes to the road and roadside infrastructure.

  2. To enhance, develop and integrate technologies for a prototype platooning system such that the defined strategies can be assessed under real world scenarios.

  3. To demonstrate how the use of platoons can lead to environmental, safety and congestion improvements.

  4. To illustrate how a new business model can be used to encourage the use of platoons with benefits to both lead vehicle operators and to platoon subscribers.



I remember an article about this in the past where they showed how much gas savings one could get as a function of the number of vehicles and the spacing. It was a fairly strong function of vehicle spacing. at bumper to bumper for 4 vehicles one could achieve 20% reduction in fuel consumption. at 1 vehicle length spacing the benefit had fallen to 10%.

I wonder how this would be implemented. For example, if I was driving by myself and saw a train could I ask permission and link up??

This seems like a viable concept. I like it.


With an all car train yes, you would need bumper-to-bumper spacing to get maximum savings. But these trains have a truck (or other large vehicle) at the front. Trucks produce such a huge wake in the air behind them the following cars can be spread out more and still get the full savings. And as an added safety feature - trucks also have a longer stopping distance than cars so if the lead driver has to do a panic stop the computers in the following cars will have an added margin for safe stopping.


Yes ai vin it adds new meaning to computer aided drafting:)


Note that fuel savings for an ICE vehicle equates to range extension for an electric vehicle.


Exactly what kind of testing did they do in that 200 km? was there any merging, panic stopping, exiting, rain, snow, ice, potholes, deer. Or did the cars just go in a straight line?

How is a non-road train car in the left lane supposed to exit to the right if there is a long line of automated cars, spaced at only 6 meters, in the right lane? If somebody nice in the road platoon lets the car cross, what happens to the platoon then? What happens if a non-convoy car just merges into the convoy (six meters is just enough for most cars) and stays in their lane? There's a million situations that aren't addressed in this article.

The big question is how do cars with different braking distances work in such a platoon. Some cars do 60-0 in 110 ft, others in 170 feet. How's that gonna work? Is the convoy only allowed to have Volvos in it?


By the way. Isn't Volvo the company that claimed their cars were safer in roll-overs, until they got caught re-enforcing the car on the inside with some kind of structure, which wasn't supposed to be visible on camera? Now they're trying to sell some other cool technology and claiming it's going to be safe, by showing some phony road testing, without providing any quantitative test results concerning safety issues.


I guess this what the experiments are designed to discover.
What happens if a driver falls asleep ?
What if we can't wake him up - does the car in front lead him to the breakdown lane and call an ambulance ?
If the cars all have radar, they should be able to see obstacles and avoid really sharp braking.
They will probably have to have ways of making the trains, so that people can work around them to get to the off ramp - also, the trains should be able to split so people can get through them.


Exactly what kind of testing did they do in that 200 km? was there any merging, panic stopping, exiting, rain, snow, ice, potholes, deer. Or did the cars just go in a straight line?

I don't know but either way this was only one in a series of tests so the problems you mention should have come up at some point: "The three-year SARTRE project has been under way since 2009. All told, the vehicles in the project have covered about 10,000 kilometers (6,214 miles)."


... also, how much of the fuel savings is eaten by the lead truck's fuel consumption?



Other studies have shown the lead vehicle in a road train also generally benefits from drag reduction because the vacuum that forms behind it gets filled by a mass that's moving with in.


The fuel consumption of the lead vehicle isn't a factor if it was going to be on the road anyway.


Fuel savings of 20%? Hard to believe and it's deceptive. If a car (Volvo) is used as the lead vehicle, and there are ten other cars in the convoy, the lead car is using as much fuel as the other ten, which adds 10%. So the 10 non-lead cars would have to use about 30% less than if all 11 cars were not in a convoy.

What's the lead car for anyway? If all the cars in the convoy are using radar, networking, vision, etc., what do you need a lead vehicle for? Apparently, it's only defining the route, for example from Baltimore to Washington. So the lead vehicle goes from Baltimore to Washington without deviating and some cars enter and exit the convoy from other roads, if they're at the right place at the right time, when the convoy goes by.

Maybe the lead vehicle could leave Baltimore every thirty minutes, while other cars line up behind it waiting for the departure time. Would people really want to do that? It only takes about 45 minutes for the trip anyway. Lots of questions about why we need car trains other than they're cool looking.


The lead vehicle with the professional driver is to provide the guidance and human judgement for the platoon. Computers aren't up to that yet.

Wind tunnel tests have found drag reductions up to 61% in following vehicles, though this is not at the separations used in the Volvo experiment.


Drag reduction is not the same thing as fuel reduction.

By using a lead vehicle, the fuel consumption is already increased by 1/(number of following vehicles). So the drag reduction would have to result in decreased fuel consumption equal to the increase caused by the lead vehicle plus 20 percent of the whole train's consumption.

My point was that it's hard to argue the road train concept based on fuel consumption, if you need a lead vehicle. However, people are pretty excited about Google's self-driving car. If that's too far in the future, then you need a "professional" driver to slam on the brakes when a deer jumps in the lane. Of course if the "professional" driver is talking on his cell phone, he may not see it. Then you would have a lot of cars totally out of control.


If the lead vehicle with professional driver is going to be on the road anyway, it is a "sunk cost".  Paying that driver out of the fuel savings in the following vehicles is just good economics.

If we could arrange to get even 30% savings in a following vehicle, savings for a train of 3 would be 20% overall, and 25% in a train of 6.  That's nothing to sneeze at.


I don't understand why the lead vehicle would be on the road anyway? It's always part of the train so you have to add that cost to the cost of the following vehicles. If there aren't any FVs, then you don't need a LV.

If you read the papers on the Sartre site, you'll see that there won't be more than about 10-15% savings in fuel because of reduced drag. That's because they won't be able to put cars closer than about 2 car lengths. That's because people involved in the simulations don't feel safe and because the cars have different stopping distances.

One paper describes how people feel comfortable passing a train to get to the next exit in a simulator. But this is hard to believe. Sure, if the train was in the right lane going a little less than the speed limit and the car is in the left lane two cars behind the LV, and needs to get in the right lane by passing the LV, then it's plausible. But what if the car in the left lane is five cars behind the LV in a 15-car train? Then it has to speed up to some illegal speed depending on the distance it needs to cover. That's not gonna work and will not pass anything like a NHTSB acceptance testing proceedure. It's another reason to have a gap big enough to allow a truck to go through, which negates fuel/drag savings.

What happens when a close-gap 15-car train is going through an interchange? Cars trying to enter from a ramp will have to stop and wait, like at a railroad crossing. Then they will have to accellerate hard, which is dangerous and uses more fuel, to merge into traffic following the train.

Car train engineers think they know what they're doing, but their plan severely underestimates the degree of complexity involved with human operation of fast-moving, dangerous machines.

What happens when a driver sees something alarming in his email, quickly decides to take manual control, and suddenly leaves the train by merging into the left lane, but a car already in the left lane? The driver panics, swerves back into the train, where the gap has already started to close, puts on his brakes, and causes the other cars to brakes suddenly. The risk of a huge pile-up, with death and severe injuries, is very real and the Sartre people have not addressed this adequately. Until they do, they shouldn't be using trial and error testing on public roads.

The comments to this entry are closed.