How maps can save connected cars from data outages

Connected cars are promising us a future of safer and less stressful driving. Driver aids built on data, knowledge and information are starting to bring this vision to life. But there’s a problem, what happens when we don’t have a data signal or cellular connection?

Today, it’s estimated around one in three vehicles on our roads in the U.S., Europe and China are equipped with a cellular data connection. That figure is only expected to grow. By 2030, 95% of all new vehicles sold will be connected, according to research from Counterpoint’s Smart Automotive service & McKinsey.

That’s all great, but connected cars rely on data networks. While coverage is getting better all the time, there are still gaps.

Take a look at netBravo, the European Commission’s crowdsourced project that gathers information about cellular connectivity data and WiFi networks. Coverage across Europe is broad, but it’s not perfect. There are plenty of areas not covered by a data connection and in most cases signal strength and speeds are poor to middling. These areas are known as “white zones.”

The EC’s netBravo uses anonymized data to plot cell data coverage across Europe. While it’s not a totally perfect depiction, it gives a good insight into the real-world conditions and usability of data networks.

Data, the backbone of connected cars

Automakers, OEMs, Tier 1s and data providers and mappers, such as TomTom, are keen to take advantage of the growing connected car market, Saurav Miglani, Automotive Product Marketing Manager at TomTom explains. In the global pursuit of making the roads safer and vehicles less stressful to drive, data brings several possibilities.

Probably the most talked about is over-the-air (OTA) updates, which allow carmakers to regularly push feature updates, like UX improvements and new apps, to their cars. In EVs, we’ve seen automakers improve the efficiency and driving range of their cars with code tweaks pushed out OTA.

Connectivity also means that vehicles can be made compliant with the latest regulations, like Intelligent Speed Assistance (ISA). Advanced driver assistance systems (ADAS) like predictive cruise control also rely on a data connection to get information about the road ahead from cloud-based location data. (If you want to know more about how this works, take a look at this article.)

Data is also crucial in EVs. Offline map data can tell us where charging stations are and help route us to them, but we can’t know if those chargers are available, or even functioning. In connected cars, real-time information about the charger status can be sent to vehicles and apps.

Then there are the more jovial things. Data connections allow us to while away the hours when we’re not driving by watching Netflix and listening to Amazon Music.

According to data gathered by TomTom on test drives around Europe, 4% of an average car journey is through white zones. Most of the time, when a car drives through a white zone it has no access to precious data that makes its driver aids and infotainment work. So, we should remember that perennial problem, what can we do when we don’t have a data connection?

Handling data in a ‘white zone’

For OTA updates and watching Netflix in your car, white zones aren’t likely to be a huge problem. But for connected vehicles packed with driver aids and automated features that rely on up-to-date data, white zones represent a problem, Miglani explains.

“Connectivity is crucial to the operations of the car; be it navigation, automated driving or ISA,” he says. “You wouldn’t like to have the unpleasant experience that results from connectivity issues or loss.”

In some cases, losing data connection could make driving more dangerous, stressful and distracting. For autonomous vehicles, even the smallest lapse in data connection can be the difference between a safely navigated trip and an accident.

On one hand, the solution would be to ensure all road networks have great cell coverage, then we’d never have to worry. Sadly, that solution is expensive to install, expensive to maintain and still doesn’t guarantee a perfect data connection. Outages, while rare, do happen and can wreak havoc. Cell network companies also need to perform maintenance, which can result in scheduled downtimes, which still need to be navigated. Network congestion can also become a significant problem, sometimes the volume of people requesting data all at the same time can turn an otherwise good connection into a white zone.

Miglani tells me there is another solution, one that can ensure vehicles have as much up-to-date data as they need, at the time they need it — and it uses maps.

Mapping connectivity

Through specialist sources, TomTom has access to data about the quality of cellular networks from more than 300-million devices worldwide. The datasets those sources generate add more than 200-billion data points every day. The data gathered measures network characteristics like upload and download speeds, latency, jitter, packet loss and time to first byte (TTFB).

Knowing this allows TomTom to track the quality of data connections and when there are network white zones. When this data is put on a map is where the real magic happens. Doing so creates a dedicated layer that shows geographically where white zones are or are likely to be. TomTom calls this the Map Connectivity Layer. It means we can physically see where we’ll possibly lose connection and where we’ll regain it, allowing us to preemptively plan for signal losses.

When we know where connection white zones are, we can send data to connected vehicles when they have a signal for the road sections where we know there won’t be a data connection.

When combined with other location technologies, such as GPS and navigation, we can adjust how data flows to a vehicle to ensure it doesn’t fall foul of a poor connection. In principle, how it works is simple. When a vehicle is approaching a known white zone, the vehicle’s connected or online services can request the data it’s going to need to make it through the white zone safely, rather than requesting it on-demand.

Take one of TomTom’s most recently released products for example, Virtual Horizon Online. It’s a cloud-based variation of the company’s Virtual Horizon product, which provides vehicle systems with a live data feed of information about the road ahead. This can be used by predictive cruise control systems to manage the drivetrain when approaching a hill. It can also share the upcoming speed limits of roads for a given route to ensure the vehicle remains ISA compliant and the driver doesn’t receive any unexpected speeding fines. But it can only do this with a data connection.

Thanks to the map connectivity layer, Virtual Horizon Online can work around connection outages by gathering all the data about the road ahead and providing it to the driver and the location tech systems that rely on it, when it still has a signal.

Miglani explains there’s a lot more going on in the background, “for Virtual Horizon Online to work, it requires a series of GPS locations from a vehicle. It then performs a process called map-matching to understand where the vehicle is and what road it’s driving on.”

“Using the latest machine learning (ML) techniques, the system calculates a most probable path (MPP). For ISA, the respective speed limits for this route will then be downloaded and stored locally for when they’re needed.”

“When the vehicle exits the white zone, it can reconnect, update its position and it will get the new speed limits for its most probable path.”

Providing vehicles with data this way is efficient and more importantly, it ensures that vehicles aren’t downloading speed limits for all roads in a white zone, only the ones that they’re likely to need.

Imagine there’s a white zone that’s five miles long, and you’re on a highway with no exits for 10 miles. The system will only download the speed limit for that one road you’re on. In this example, the MPP will have to be along the same road for at least 10 miles. So, there’s no need to download the speed limits for the roads in a town that’s only two miles away but can’t be accessed from your current heading.

A meaningful data-driven map

When we think of maps, most of the time we focus on things like where places are, businesses, your friend’s house. Or we worry about how to get to places, what turnings to take and where to go. But as Miglani explains to me, maps are more than that and depending on the data we meld with them, we can gain new contexts of meaning and information.

In this case, we’re able to map signal outages and elevate the effectiveness of navigation, routing and advanced driver assistance systems. For road safety and driving ease, it can only be a good thing.