Self-driving cars and trucks fire the imagination. Besides offering a more relaxed journey for the driver, they hold the promise of reduced traffic congestion, lower CO2 emissions, and safer traffic. At the same time, the media are asking for a realistic view of what happens when computers do the driving task for us. What is the potential of self-driving vehicles?
Today’s trends in traffic ask for a change: one only has to think of ever-increasing congestion, rising accident rates, and the need to substantially reduce CO2 emissions. However, some important steps will need be taken before the necessary transition to self-driving cars vehicles be made efficiently, safely, sustainably, and with broad social acceptance and support.
Taking these steps will require cooperation between vehicle manufacturers, suppliers, knowledge institutes, and the government. Towards this end, TNO is working with the vehicle industry in national and international projects on technological development and validation; the human factor in automated driving; ICT; and socio-economic impact assessment. These four preconditions – which are also interrelated – are required to make the desired transition possible.
Technological development and validation
“Lots of experiments (physical and virtual) are being performed with vehicles that use new automated driving functions, from the Uber-car to the Tesla,” says Automated Driving programme manager Maurice Kwakkernaat of TNO. “For instance, before a new vehicle is allowed to drive on the road, it has to be tested under circumstances it might encounter, in order to be certain that its automated systems (steering, braking, decision making, interaction with other traffic) function properly.”
“People tend to brake strongly and then accelerate again. If you can automate that, you can reduce those sort of dynamics, improving safety and strongly reducing CO2 emissions”
A particularly promising area is the combination of automated driving and information exchange between vehicles, infrastructure and other road users. This is because this technology can deliver advantages even when not all road users are making use of automated driving. “People have a tendency to overreact to situations; they brake strongly and then accelerate again, for example. If you can automate that, and make use of the intentions of those driving in front of you through communications networks, you can predict driving behaviour so as to reduce those sort of dynamics, improving safety and significantly reducing CO2 emissions. That can’t be done with automation alone.”
“We are also working on detecting objects, pedestrians, or other situations causing a safety risk”
Can this transition not be accelerated? “It’s a complex process,” explains Kwakkernaat. “Besides technological development and validation, you need to have the right ecosystem of a variety of public and private parties. With projects on platooning, for instance, virtual validation, and government programmes we are ensuring that the potential of self-driving vehicles can actually be utilized – safely, and in such a way that it benefits both the individual and society as a whole.”
The human factor
Marieke Martens is senior scientist at TNO and professor of Intelligent Transport Systems and Human Factors at the University of Twente. “One driver may happily hand over control of the pedals and the steering, thereby overestimating the car’s intelligence; while another driver may fail to understand his/her car altogether, thereby underestimating its self-driving functions so severely that (s)he prefers to switch all these functions off. The more we understand drivers and passengers in a self-driving car, the better we can identify how such a vehicle should behave and how it should communicate.”
“The more we understand about drivers and passengers in a self-driving car, the better we can identify how such a system should behave”
To learn as much as possible about driver behaviour TNO not only carries out driving simulator studies of future systems, but has just equipped – on behalf of the Ministry of Infrastructure and Water Management; the Dutch national road authority Rijkswaterstaat; and the Netherlands Vehicle Authorities – nine current-generation vehicles with self-driving functions with numerous cameras and additional sensors and loggings.
Martens explains: “We continuously measured and logged everythings that happened. How do other road users respond, , where does the driver put his hands and feet, does the driver keep his/her eyes n the road, and so on. We can even see how their learning curve developed. We just began carrying out the first analyses of this data; we will have the first results by the end of the year.” This information can be used to improve future systems.
The utilization and exchange of an ever-growing arsenal of external information sources could also make an important contribution towards automated driving. “Of course, this information has to be reliable and up to date, and mobile communication and internet technologies have to be ready for it,” notes project manager Bram van den Ende of TNO. “How do you ensure that information reaches the right vehicle, that the ICT network is not overloaded, and that this information can be exchanged securely?”
“How do you ensure that information reaches the right vehicle, that the network is not overloaded, and that this information can be exchanged securely?”
In a European project, AUTOPILOT, TNO has invited more than 10 parties in the ‘Brainport region Eindhoven-Helmond’ to show what could be done with the Internet of Things and pre-5G technologies. Van den Ende: “This pilot project is making a substantial contribution towards international standards, for example on the language employed in communications between the vehicle and the cloud, to allow platooning with different vehicles.”
The European project Ensemble, coordinated by TNO, is concentrating on collaboration in the secure, reliable exchange of information between different parties. “ICT and information exchange can ensure that automated vehicles get much smarter, but it does make demands on the communication infrastructure. 5G networks will play an important role in this, as well as new possibilities with satellite communications.”
Identifying and assessing the effects
The researcher Isabel Wilmink of TNO is working to identify the effects of these technologies on traffic and transport choices, and the ensuing development of policy. “If automated electric taxis are cheap because they need no drivers, will people leave their bikes at home and take a taxi instead? What will traffic flows look like if they are composed of both automated and manually driven vehicles? Are there already benefits, in terms of traffic safety and emissions? And with the rise of car-sharing concepts which mean that fewer people will own a car, and ‘Mobility as a Service’ concepts, what role are automated vehicles going to play?”
Wilmink has more questions that need answers: “Could traffic lanes be narrower? Would we still need a hard shoulder? Or would we need to build even more lanes, because people will travel longer distances? If people can work while they drive, would they be willing to live further away from their workplace? Using interviews, pilots on test tracks and public roads, driving simulator research, and traffic and transport models we are working on answering these kinds of questions as accurately as possible, so that we can get the most out of the potential of the self-driving car.”
The self-driving car has potential – this much is certain. But there is still some way to go. Putting this type of vehicle on the road demands close cooperation between TNO, the industry, and government. We are eager to enter that cooperation. Do you have ideas of your own, or would you like to know more? We look forward to hearing from you.
More information can be found on the page 'TNO and the self-driving car'. TNO outlines the potential of the self-driving car. Different facets are discussed: technology development and validation, effects on traffic flow, ICT and the human factor. You will also find relevant links to pages on TNO.nl.
- Traffic & Transport