No 'smart traffic' without your vehicle and location data

This article in four points:

• If it is up to the European Commission, smart, data-driven technological applications will lead to major improvements in road traffic within a few decades. In recent years, concerning smart mobility many hundreds of trials conducted at home and abroad. The Netherlands is among the leaders in Europe in this field.
• Smart mobility applications address two, overlapping issues: road safety and traffic efficiency. Traffic and transport are now inextricably linked to data traffic. Without large amounts of data from vehicles and mobile phones hitching a ride, road authorities such as Rijkswaterstaat would be much less able to provide road users with up-to-date traffic information. Or to maintain roads in a well-informed and cost-effective way.
• Most of the time, motorists do not benefit from their own vehicle data because others take advantage of it. In today's smart mobility applications, this is different as the data takes a circular path: via commercial parties and government agencies, it comes back to the road user in the form of customised services. Whereas the providers of navigation services know exactly where you are driving to, and the same applies to car manufacturers if you allow data sharing with them, the government cannot deduce that from their aggregated and anonymised datasets.
• A number of smart mobility applications have already been put into practice, while others are still just future. Expectations are still highest regarding semi-autonomous vehicles that automatically adapt to each other, but Cooperative Intelligent Transport Systems are hardly getting off the ground as yet, however.

There are few parts of the Internet of Things where expectations are as high as for smart mobility. Imagine vehicles with a high degree of autonomy that adapt to each other's position and speed, optimal traffic flow that puts an end to the congestion problem, detected danger on the road that translates into an alert on the dashboard within seconds, only a minuscule number of traffic accidents and a drastic reduction in CO2 emissions from cars. It sounds almost too good to be true but who wouldn't sign up for it? The European Commission has already done just that. It firmly believes that smart, data-driven technological applications in road traffic will lead to huge improvements within a few decades.

Smart mobility sounds hip, but the road towards it is long, bumpy and uphill. It is a very large and tough dossier that requires a lot of coordination between EU countries and takes years to reach agreements on technical standards, on the use and sharing of data and, for example, on motorists' privacy.

Smart mobility is partly already a reality, but in many respects it has yet to take off. Some countries are also (much) further along with this than others. The Netherlands is among the leaders in Europe. Up to 2018 alone, a total of 445 trials and pilots were conducted here that focused on smart traffic solutions. They did not sit still in the following years either, as evidenced by the many dozens of Rijkswaterstaat projects that have just been completed or are still ongoing.

As part of our article series on connected cars in two parts, we highlight some forms of smart mobility in which vehicle data play an essential role. Showing who vehicle data ends up with and what happens to it afterwards is one of the starting points of this series. Earlier, we followed the vehicle data trail to car manufacturers, vehicle data hubs and fleet managers, the automotive aftermarket, hackers, Big Tech companies and information- and investigation services. Road authorities and related parties are now added to this list.

Although motorists at least on paper Be in charge of their vehicle data, others often use them in practice, without motorists themselves benefiting. In smart mobility applications that have already been put into practice, the situation is clearly different. In these, data follows a circular path: in the form of traffic information and customised services, it is given a second life and returned to the road user. Yet there is something bittersweet about this because you sometimes pay for it twice: with your data anyway and - if a particular service comes with a price tag - also with your wallet.

From location data to traffic information

Smart mobility broadly relates to two, overlapping issues: road safety and traffic efficiency. Where in particular Probe Vehicle Data relevant to that first, are Floating vehicle data indispensable for that second. In a previous article we previously explained that Probe Vehicle Data is data actually generated by a car. Floating vehicle data on the other hand, come from devices that merely 'hitch a ride' in a car: notably mobile phones but also boxes and dongles installed in a vehicle 'afterwards' for trip and driving style recording. It is not real vehicle data, mainly time-stamped location data.

Motorists use navigation apps such as Apple Maps, Google Maps, Waze (worth over a billion dollars last year acquired by Google) and Flitsmeister (Be-Mobile), as well as from built-in tracking devices of, for example, insurers, that the resulting data form a very representative and valuable picture of the situation and traffic intensity on the road. Inherent in navigation services is that the providers know exactly where you drive. There is no getting around this, even if you use offline maps.

If you allow these companies to direct you to your destination, you agree that your location data will be sold in real time to parties such as the National Road Traffic Data Portal (NDW). The NDW is a partnership between the state, the 12 provinces and several transport regions. According to the specific needs of road authorities, NDW purchases the data in anonymised and aggregated form for its partners, but also makes it openly available on a Open Data Portal. Among others, ANWB uses it for its traffic information service to supplement the location data of everyone driving around with ANWB driving insurance (location data that ANWB itself also sells to road authorities, mostly municipalities - at least that way the union recoups the discount policyholders get for their safe driving).

How exactly Google & Co anonymise and aggregate their location data, they never clarify. It is known that aggregated data (in the wrong hands) can sometimes still be pulled apart in such a way that they can still be traced back to individuals. But from the real-time datasets NDW puts online, it is in any case not possible to deduce who drives where (although the government has other methods for). How busy it is somewhere, though. Partly thanks to Floating Vehicle Data, there is now a much better view of traffic throughout the day in the whole of the Netherlands, not just in the Randstad, as used to be the case. Even provincial roads are now in view.

This allows road users to be better informed about traffic conditions, for example via matrix signs above motorways. Information appearing on such signs used to be based on data from relatively expensive detection loops over which cars drive. Loops that never gave a complete picture as they were installed on only half of the road network. Today, matrix signs almost entirely fed With messages based on Floating Vehicle Data.

Traffic data not only enable improve flow, they also make it easier in a general sense to implement traffic policies, think for example of addressing cumbersome cut-through traffic. Traffic problems will never be solved based on data alone, and blind faith in (Big) data is not to be recommended anyway. Yet it undeniably provides governments and road authorities with valuable insights. Commercial parties such as TomTom and Lord show that too: on a global scale they know exactly what traffic patterns look like, where you can drive on unconcerned and in which cities you are hopelessly stalled every day.

Car manufacturers must compulsorily hand over data

Smart mobility applications designed to immediately promote road safety are coming in a different way. In 2013, the EU required car manufacturers to share certain vehicle data (Probe Vehicle Data) in aggregated form with the NDWs in Europe in case of an accident, a slippery or bad road surface or poor visibility, for example. Based on that 'minimum universal road safety information', NDWs can in turn notify motorists of an unsafe situation as quickly as possible. That happens within seconds, which is more than enough. Motorists who see a problem ahead 'five seconds away' need not be alerted anyway. If you see mischief through your windscreen, you should take immediate action, it should not depend on an alert you get through your an app or from your car. The point is precisely to warn the traffic behind.

The regulation in question (886/2013) was rather ahead of reality, as connected cars hardly existed at the time. Although they have since slowly made their appearance, manufacturers were stubborn for years: they were hardly willing to release their Safety Related Traffic Information (SRTI). They wanted to keep this to themselves because they saw a new business model emerging around vehicle data and wanted to decide for themselves. Moreover, they immediately realised that enforcement in this area was not much of a concern.

Under the regulation, it is up to EU countries, not the European Commission, to ensure that manufacturers and related parties fulfil their obligation and actually share their SRTI data. To this end, when the legislation came into force, the Netherlands sent over 330 letters reminding all kinds of parties of this duty. But that was all well and good, and elsewhere enforcement did not amount to much. EU countries then asked the European Commission: what now? After which there was silence for a long time. Until about five years ago, when the Commission made it clear to car manufacturers that road safety was at stake and that SRTI data really had to be shared.

Since then, significant steps have been taken. Within the European partnership 'Data For Road Safety' warnings do now become shared between manufacturers, road authorities and other relevant service providers. As it would also make very little sense to summon manufacturers who might still be in default in each EU country separately to the penalty box, EU countries within the project are trying to National Access Point Coordination Organisation for Europe (NAPCORE) to harmonise and jointly enforce matters in this area.

Hole in the road

And so in practice, the Department of Public Works gathers information on road conditions through NDW using data from hundreds of thousands of connected cars. If many cars report that in one particular location the shock absorbers are bulging, there is almost certainly a pothole or hole in the road surface. Is there massive braking where it usually never happens? Then there is most likely something on the road. No lane assist for some time? Possibly the lines are out of order. Does the ABS kick in with one car after another? Perhaps there is oil on the asphalt or the road surface is worn.

That certainty is best done by the Department of Public Works itself by constantly traversing the country with two expensive ARAN (Automatic Road Analyzers) cars equipped with cameras and highly sensitive measuring equipment. After those cars went up in flames in a fire in 2019 and there were temporarily no replacements, the (slightly less accurate) vehicle data from huge numbers of road users provided answers in a way that was previously unheard of.

If you are very concerned about your privacy and don't even want your car manufacturer to share SRTI data from your car with governments, you should change your car's privacy settings so that no more data is transmitted to the manufacturer. The main question, however, is to what extent this is really possible. For example, the mere emergency call system eCall in principle, cannot just be switched off.

Accurate information instead of forecasts

As a follow-up to regulation 886, multi-year trials are taking place on Dutch roads for road safety purposes. A major project in the Netherlands that recently came to an end after three years is Road Monitor (RoMo) of the Ministry of Transport and procurement partner Mercedes. There, too, it was primarily about mapping the state of roads in real time, and adjusting road maintenance accordingly.

Successfully, because the test showed that maintenance can be planned much more dynamically and accurately. This is also much more cost-effective than working on the basis of one-yearly measurements of asphalt and line wear, and the resulting prognoses. Those prognoses, while not out of the blue, are at most educated guesses. They can differ considerably from reality due to various factors (heavier trucks, different traffic intensity, weather and climate change, etc.). For a long time, all kinds of renovation and replacement programmes existed without the Department of Public Works really knowing the quality of the roads. That time is coming to an end.

In a nutshell RoMo worked as follows: Every Mercedes owner in the Netherlands was asked at some point for entirely voluntary cooperation. Most were not interested, but several tens of thousands were willing to share their vehicle data. The car manufacturer aggregated the data from the relevant cars (it remains unclear on the basis of which technology) and distilled the desired information from it for the ministry ('use cases'). Mercedes is one of the few car manufacturers that - apart from vehicle data hubs to - can enrich sensor data themselves and convert it into a desired information product.

On a dashboard built by Mercedes, the ministry could see almost instantly where on the road something was going on. For example, at location Y, X number of cars noted that it had become slippery because the wheels were slipping. Combined with current weather information, you could then conclude that there must be black ice. Who had driven where at what time could not be deduced from anything by the ministry or the road authorities. The Personal Data Authority, which had been consulted for RoMo, was very appreciative of this approach.

RoMo will in the coming years continue, except that with RoMo II, there will be multiple procurement partners. Indeed, the larger the number of car manufacturers participating, the more representative the picture of the road situation.

Another, currently ongoing project is 'Safety Priority Services'. In it, users of ANWB, Be-Mobile, Hyundai, Inrix, Kia and TomTom will receive an alert on the dashboard within seconds when an emergency service with flashing light and siren approaches. Road users - whose data will be anonymised in the trial - must first consent to participate.

'Cooperative driving' still future for now

The smart mobility applications discussed so far have proven themselves and are getting better and better. The same cannot be said for now about Cooperative Intelligent Transport Systems (C-ITS). Here, vehicle data does not first pass over several disks and through several parties to later reach motorists again, but cars communicate directly with each other about danger on the road, distances, speeds and vehicle dimensions (Vehicle-to-Vehicle, V2V). Or they communicate with roadside infrastructure such as smart traffic lights (Vehicle-to-Everything, V2X).

That sounds futuristic and will remain so for some time yet because C-ITS have hardly been put into practice on a really large scale. There are several reasons for this. These include the fact that the European privacy regulator drew a blank line on this years ago due to inadequate privacy safeguards; it is not yet a foregone conclusion which technology should be used for cooperative driving (Wi-Fi or 5G); most cars are not yet autonomous enough to take part in this; and especially the European Commission is actually much more enthusiastic about cooperative driving than the car industry itself, which does not immediately see an attractive business model in it. We will explain exactly what all this is about in the second part on smart mobility.

A summary of this article has been published at PONT Data & Privacy, see No 'smart traffic' without your vehicle and location data - PONT Data&Privacy (privacy-web.co.uk).