Autonomous truck convoys: the question is when, not if

The grand vision of Europe’s freight transport is of sleek convoys of autonomous trucks travelling along thousands of miles of ‘smart’ roads. Siegfried Mortkowitz reports on what it will take to realise it

A streamlined convoy of autonomous trucks travelling along thousands of miles of ‘smart’ roads, with the driver’s function reduced to maintenance and security, is the grand vision of Europe’s freight transport.
    
“I would say we will eventually see part of long-haul transport on roads that are ‘green corridors,’” says Fredrik Callenryd, senior business analyst for truck manufacturer Scania. “So that we would see autonomous trucks travelling in road trains in the foreseeable future, with drivers navigating only when they leave these green corridors, to travel to the depot, for example.”
    
European Union legislation due to come into effect is likely to speed the realisation of this vision, according to Gareth Owen, principal analyst at ABI Research.

The directive, which takes effect in November 2015, calls for all newly introduced models of trucks 3.5 tons and over to be equipped with both advanced emergency braking systems (AEBS) and lane departure warning systems (LDWS).

AEBS uses sensors to alert the driver to an upcoming obstacle, such as a traffic jam or when his vehicle is coming too close to a vehicle in front of him. If the driver does not react in time, the system automatically triggers a braking response that either prevents a collision or reduces the force of the impact. LDWS alerts the driver when the vehicle begins to drift out of the lane.

The road to driverless trucks
Both AEBS and LDWS are integral elements of advanced driver assistance systems (ADAS) that are slowly taking over the truck driver’s decision‑making duties, and which – Callenryd, Owen and other trade insiders say – are key building blocks of what ultimately will become the autonomous, or self-driven, commercial vehicle.

Already, autonomous vehicles are in use in open-pit mining and at some airports. Self‑driven mining trucks are, for example, able to negotiate roads within mining sites as long as they travel along fixed routes. And driverless transit systems, which travel on rails or guide ways, are in use at some major airports, such as London’s Heathrow and Orly in Paris.
    
Many long-haul trucks are already using cruise control systems similar to what is standard in passenger planes.

According to Callenryd, there are two models of these systems: the basic system, in which the driver sets the speed of the vehicle, and the system maintains it, and the advanced version, in which the vehicle takes into account uphill and downhill sections and adapts to run at the optimum fuel-saving speeds. “These advanced systems have already proven to be more efficient than our best drivers in tests,” Callenryd says.
    
Scania’s new solution is called Eco-roll, and it uses both GPS and topographic maps to calculate whether cruising in neutral down a hill or using engine braking with the fuel supply switched off is best for the vehicle’s kinetic energy. Eco-roll is set to come standard for long-haulage trucks in regions where topographic data is available.

Volvo’s take on AEBS and LDWS
Most truck manufacturers already offer a version of AEBS and LDWS. Claes Avedal, traffic safety manager at Volvo Group Trucks Technology, says Volvo has been offering LDWS on its FH and FM models since 2007, and AEBS on the FH since 2012.
    
Volvo’s lane departure warning system uses an optical camera sensor to detect when a vehicle begins to drift out of its lane and audible and visual warnings to alert the driver. However, the technology is not yet foolproof. “Functionality is dependent on visible lane markings,” Avedal says. Fog, rain and darkness are usually no problem. But if there are no lane markings, the system will not work. Snow and ice on the road which cover the lane markings are also problematic.

For AEBS, Volvo uses both camera and radar sensors to detect traffic ahead. “Radar is used to detect traffic at longer distances, and the camera is used for object detection at shorter distances ahead of the truck,” Avedal says. “By combining camera and radar technology we can detect more traffic scenarios, like stationary vehicles, and avoid false warnings.“ The initial visual AEBS warning is projected on the windscreen in the driver’s forward field of vision so that the driver will react quickly.
    
In addition, Avedal says, Volvo trucks are equipped with several other ADAS features, such as adaptive cruise control (ACC), which automatically adjusts the vehicle’s speed to maintain a safe distance between it and vehicles and other obstacles in front of it; lane change support (LCS), which helps the driver when changing lanes toward the passenger side; and driver alert support (DAS), which alerts the driver when he is getting drowsy or nodding off.
    
The next step for Volvo is a technology that enables vehicles to communicate and exchange information with other vehicles (V2V) and with the infrastructure (V2I), known as Cooperative Intelligent Transportation Systems, or C-ITS. “One example of this is to provide earlier warnings to the driver of traffic conditions ahead, such as reduced visibility, slippery road, accidents or other obstacles,” Avedal says.
    
In addition to OEMs, manufacturers of ADAS such as Delphi are working to perfect the technology while offering aftermarket products to help fleets meet the new EU regulation. Delphi, for example, offers – among other ADAS products – a radar sensor that detects moving and stationary objects at both mid- and long-range distances.

Not a question of if, but when
Avedal, Owen and Callenryd all agree that it is not a question of if, but of when autonomous truck trains become a reality, and all three say that the benefits are potentially very significant. “We believe that a step-by-step introduction of more autonomous functions will improve traffic safety, reduce congestions and fuel consumption,” Avedal says.
    
“If you have a convoy of trucks with just one person in the lead truck and ten driverless vehicles behind it, you have significant cost savings,” Owen says. “First of all, you save on fuel. Trials have shown that these convoys can save up to 15 per cent on fuel, in the same way bicycle racers save energy by riding in a single file, to reduce wind resistance.”
    
There would be savings because there are far fewer drivers to pay. And traffic safety would also improve.

“Human beings are lousy drivers,” Callenryd says. “How many people died on European roads last year, 40,000?  If we allowed a system to kill 40,000 people a year, then we could probably implement a fully autonomous system today. But we don’t allow machines to be as inefficient as humans.”
    
“Some 90 per cent of accidents are caused by drivers,” Owen says. “That’s a very big target to hit.”

Technical and legal obstacles
Obviously, many challenges remain to be met, most of them technical.
    
“The quality of real-time traffic information needs to be improved,” Callenryd says. But he adds that there have been “great improvements” in the technology over the past five years, and he expects that process to accelerate. Owen underlines the current high cost of radar sensors that are used to detect traffic and other obstacles. “But I expect a significant drop in the price of radar sensors,” he says. “They will eventually be mass market, not niche.”
    
However, legal obstacles may actually be more difficult to overcome.
    
According to Callenryd, cross-border road traffic in Europe is regulated by the Vienna Convention, agreed in 1968. “it is very clear that drivers should be in control of vehicles,” he says. “There are no legal policies for computer-driven vehicles. We need a new legal framework that resolves such questions as how we judge between state‑of-the-art vehicles and ‘dumb’ cars.” The problem, Callenryd says, is one of liability. “If an autonomous car causes an accident, who is responsible? The car industry is not ready to take this responsibility.”
    
Owen agrees. “In case of an accident, the blame will probably shift more to the OEMs. OEMs will not release [autonomous] vehicles until they’re sure the liability is containable,” he says. He sees the introduction of what he calls semi-autonomous commercial vehicles – in which the driver can take over some functions – by about 2020. “Truck trains may be possible by the year 2025,” he says. “It’s a gradual process.”

This article was first published on Telematics Update