In the Spotlight: Naturalistic Driving Studies

Rob Eenink, consortium leader of PROLOGUE – the EU FP7 feasibility study for a large scale naturalistic driving project, and Dr. Kenneth L. Campbell, Chief Programme Officer for the second Strategic Highway Research Programme (SHRP 2), share their insights on naturalistic driving studies on both sides of the Atlantic.

What is the purpose of naturalistic driving studies (NDS)?

 

Kenneth L. Campbell: The primary strength of studies such as PROLOGUE or SHRP 2 is that objective information on driver behaviour, and the context that behaviour occurs in, will be collected on a large sample of volunteer drivers during the normal use of their vehicles. 
Rob Eenink: The purpose of NDS is to know how people behave in a natural way in their own car. With traditional research methods there are, for instance, instrumented cars that are driven for say a half hour or an hour with someone sitting next to the driver, with computers that are visible… In this case it is not the driver’s day-to-day behaviour that is analysed, as opposed to NDS.

How can you assure that monitoring devices do not influence test drivers?

R. Eenink: First of all, you do so by ensuring that the equipment is small enough and that drivers don’t have to do anything out of the ordinary to start their car. This is possible nowadays because Information and Communication Technologies have allowed for the development of small enough sensors, cameras, data strorage…

K. Campbell: What’s more, previous naturalistic driving studies have demonstrated that drivers seem to quickly forget they are being monitored. It may be true that some of the worst drivers or driving behaviours are self-censored, but prior studies show that there is no shortage of high-risk behaviour to study.

R. Eenink: That’s right. And I believe that once you see the type of behaviour that people wouldn’t want to be displayed publically, you hence can assume that they are not aware all the time of their being observed. I believe that in America this is called the ‘nose-picking factor’.

What is the added value of naturalistic driving studies?

K. Campbell: Traffic safety is now focused on preventing collisions in addition to the traditional focus on minimising injury in the event of a collision. The greatest need for collision prevention is information on the role of the driver. Collision investigations yield little objective information on what drivers did leading up to a collision. Only by collecting objective information during normal driving as well as situations leading to collisions can we identify the combination of circumstances and driver actions that increase collision risk.

R. Eenink: So if you know how people drive ‘naturally’ you can identify what they are able to do and where they would need support. For example, older drivers are generally more experienced, so we can assume they are safe drivers with the exception of situations where they need to take a lot of decisions in a very short time, for instance when they have to turn left – that is if you drive on the right side of the road. Based on the information that is collected, we could imagine creating or adapting advanced driver assistance systems (ADAS) that would help minimise the risk linked to that type of manoeuvre.

Are age-related differences taken into account by PROLOGUE and SHRP 2 then?

K. Campbell: The expected growth of older drivers (from an estimated 15% in 2010 to approximately 25% by 2030 in the US) is a specific focus of the SHRP 2 NDS and we are attempting to over-sample this group, as well as the younger drivers, in order to be able to address age-related differences in driving and collision risk.

R. Eenink: PROLOGUE didn’t have a specific pilot on older drivers. However, we had a couple of pilots that addressed young drivers. In Austria they looked into the effects of driver training, and in Israel they focused on peer influence among young drivers. This is also an added value of NDS: because you can see how people actually drive, you can pinpoint the differences between age groups and think of measures that are specific for certain types of road users.

What is the general impact of naturalistic driving studies?

K. Campbell: The goal of the SHRP 2 NDS was to address high-priority safety issues in the U.S. such as road departure, intersection safety and the role of driver behaviour. However, since the data acquisition systems (DAS) will be recording all driving, there are few limits to the potential benefits. Roadway characteristics will also be collected to support study of how the roadway influences driver behaviour and the effectiveness of roadway countermeasures. Certainly auto manufacturers will be very interested in how drivers use existing technology. Driver behaviour programs will benefit from more accurate information on the role of inattention, fatigue and impairment. All of the results have possible implications for policy decisions.

R. Eenink: Indeed, the knowledge gathered from NDS is of interest to anyone involved in road safety, environmental issues or traffic management. This includes the car manufacturers, insurers, but also public authorities – road authorities for instance can attune their policy to the results of the study. The impact of NDS goes beyond road safety. With naturalistic driving one can assess the effectiveness of eco-driving programmes, for example, and make them more effective or even envisage changing car and road design to use less fuel and reduce carbon dioxide emissions. PROLOGUE finished on 31st July and its main aim was twofold: on the one hand we wanted to give recommendations for a large-scale European study, and on the other hand we wanted to get support for such a study. There is now a topic for an Integrated Project in Europe: a large scale study that will follow PROLOGUE’s 11 recommendations.

What are the main differences and similarities between Europe and the United States when it comes to NDS?

R. Eenink: First of all, I believe the main difference is the magnitude of the project in the U.S., not only in terms of size but also in terms of budget. I would add that the Americans are a little further ahead of us with respect to R&D and the equipment itself. In Europe, we usually have different sources of equipment, different functionalities, different prices, whereas in the United States they have agreed upon using certain types of equipment only. I think there is a lot of added value in using one type of naturalistic driving equipment, because you get comparable databases and it facilitates the analysis. Speaking of which, in America they are trying to build a very large database which will be available to anyone interested in specific research questions. In Europe however, we focus on a few items, for instance eco-driving and specific risk factors, and in parallel a large database exists in which one can try and find the answers to questions we didn’t foresee.

K. Campbell: As with the first Strategic Highway Research Program, it is anticipated the Congress will fund a SHRP 2 Implementation phase in the next highway bill to support the transition of research results into highway safety practice. We anticipate that the SHRP 2 implementation phase will indeed include funding for a data steward to maintain the NDS data and provide access to researchers for years to come. As for the similarities between the two continents, I would say that prior to SHRP 2, the U.S. and Europe had been more focused on Field Operational Tests (FOT). FOTs use very similar instrumentation but are focused on the evaluation of specific (usually vehicle-based) technology and generally have much smaller numbers of participants. Now the U.S., Europe and other countries are moving to the broader NDS to fill an important need to support development of future safety systems.

Is a global strategic alliance on NDS foreseeable in the future?

K. Campbell: Adopting common methods, data sharing and the ability to make cross-country comparisons would greatly aid highway safety progress. It is certainly foreseeable.

R. Eenink: Definitely. And I think that is where we are headed: we have very good contacts with Australia and Japan, SHRP 2 had representation in PROLOGUE Advisory Board, and we have representation within SHRP 2 as well. Overall the most interesting situation within NDS is if you can observe a crash, which is a very rare occurrence. If we cooperate at transatlantic or even at global level, we can observe more of these events and better analyse them. That would certainly be a plus.