How AI will help the 5G connected vehicles of the future
As telecom operators look for additional ways to unlock new revenues, smart cities are currently being actively promoted, and monetization of data is emerging as a prominent—if potentially controversial—revenue source.
December 18, 2018
Empowering connected and autonomous vehicles is being promoted as one of the key use cases for 5G. But anyone who regularly travels will know that even 3G and 4G coverage is still patchy on many routes. The great ambition of covering every square meter of a country’s road network with 5G—and keeping it covered without temporary service drops—is practically unachievable.
So how can we manage this less-than-perfect connectivity and still capitalize on the promised benefits of the connected driving experience? First of all let us look at the range of features that will be empowered by, and thus dependent on, connectivity. Aside from regular voice communications, the best known connected feature is infotainment—streaming video (for the passengers), music and spoken audio, satellite navigation, and traffic alerts. Infotainment and navigation systems currently only receive information on road and traffic conditions. The complexity of this data is set to increase rapidly as large numbers of vehicles become capable of uploading data to the cloud about parameters such as average traffic speed, congestion, hold-ups and diversions, and weather conditions. The data flow will then become two-way, with vehicles both sending and receiving actionable data that has been crowd-sourced from other vehicles along the route.
Another particularly important feature is the collection of telemetry data on vehicle performance and system status, which can be uploaded to the cloud and accessed by the vehicle dealerships. This data could be used to personalize service and maintenance schedules, and also to gain insights into aggregated vehicle performance and driver behavior across their entire customer base. It also facilitates personalized safety alerts, for example if a particular subsystem or metric goes out of specification and might be likely to cause a breakdown or even an accident.
With vehicles becoming highly dependent on near-perfect connectivity, there needs to be a way to allow these features still to function when the connection is degraded for some reason. The answer is to use Artificial Intelligence AI to build a real-time picture of the connection quality over the entire route the vehicle is traveling, and any likely routes it could take. By predicting where and when the connectivity may be compromised, it can decide to cache enough media data beforehand so that the user experience is not adversely affected. And in the case of the vehicle telemetry data, it can arrange for uploads or ‘flushes’ to occur when the connection is good, and to store data on board when it isn’t so that it can be uploaded later. At the same time, the data that is being gathered can be used to inform a change of route, either to stay better connected or to avoid areas of congestion.
This AI-powered proactive connectivity prediction will also be invaluable to commercial and haulage fleet managers who rely on a good connection to monitor and communicate with their vehicles. Over-the-air fault detection and repair will both improve safety and reduce maintenance down-time. In addition, the available data will simplify detection of fraudulent driver activities or attempted hijack.
The ultimate possibilities of connected and autonomous vehicles are only just beginning to take shape, but we can be sure that AI and continuous monitoring of connection quality will play a big part in making these concepts a reality.
- Read more about smart mobility and connected mobility
- Read more about driver experience mobility
- Read more about connected car technology