Connecting the Connected Car

The way to do this is to break down the world into octagons and then to characterize the coverage in each one.  The best way to do that is to use the TCUs that are in the field to collect that data.  To minimize cost, sampling and exception techniques need to be used.  Building these maps may be more complicated than it seems at first because of how global roaming is done today.

Cellular networks tend to follow national boundaries.  So a vehicle in Europe may, on an average day, roam across two other national networks in addition to its home network.  In the U.S. on an average day, most users only roam when they encounter a coverage hole in their contracted operator’s network and another operator’s network has coverage.  Other parts of the world lie someplace between these two extremes.

Cellular networks have a way to assign priorities to classes of customers.  Operators consider their own contracted customers who they have competed to get and keep (most even track customer acquisition cost) as having the highest priority.  In some countries/networks, first responders and other critical government users may also have very high priority.  On the other hand, another network’s customer roaming on their network is not treated to much of any priority.  For example, the CTO of a major European operator with the most expensive plan available on his home network, vacationing in a nearby country could not reliably receive text messages nor access the web.  His wife, using the cheapest one-time phone purchased at a local drug store in the vacation country, was able to watch movies on her phone, etc. 

Furthermore, an operator with a customer roaming on another operator’s network can’t send a trouble ticket to the operator supporting the roamer.  So when there is trouble with roaming, it is hard for the network hosting roamers to detect and fix the problem.

These roaming problems face an automotive manufacturer seeking to sell a standard product that will work anywhere in the world.  The auto OEM needs reliable low latency communication to make the product provide the best service in today’s context and maybe as soon as next year to effectively operate in autonomous mode.  It is not practical for the OEM to have contracts with a wide profusion of Cellco’s all over the world.  They need to be able to have a contract with one (or two/three for redundancy) and roam on the networks in parts of the world where it is practical to use automotive products (not all parts of the world make sense today – for example Mt. Everest).

Given the current situation, there are three steps that can be taken quickly.  The first is for operators that want to service auto OEMs to start mapping connectivity.  Please note that we say “connectivity” not “coverage”.  That is, what users roaming from the auto OEM’s contracted network “see” while roaming on other networks around the world.

Second, traditional roaming contracts need to be updated to include a higher priority roaming service capability for these auto users.  These higher priority SLA’s (Service Level Agreements) may include expectations (and/or restrictions) on the size of the messages exchanged and the frequency of messages, etc. to manage the load on the roaming network.

Finally, these updated contracts should include the ability for the contracted network to send the roaming network trouble tickets.  This way, problems that roaming vehicles face can be quickly identified and resolved.

These three steps can go a long way to meeting the immediate needs of auto OEMs without incurring great expense, while opening up a new revenue stream for Cellco’s.  On this foundation the additional vehicle-to-vehicle, vehicle-to-road communications services, orchestration capability, and privacy / security controls can be built.  The danger is that if Cellco’s don’t begin rapidly to implement these steps, other less effective work-arounds will be adopted and Cellco’s will face limited access to this valuable market.