One of the main themes at the Connected Infotainment event in London this week, was exactly how cars will get themselves connected to the internet?
There was wide agreement that within 10 years most new cars will be in direct wireless contact with each other, and the outside world, to enable a range of cloud-based services. These will include routine updates on traffic, weather, and road conditions ahead, as well as more personal services such as live music and video streaming, concierge support, real-time vehicle diagnostics and even software functionality updates to the vehicle over the airwaves. The concepts of remote repairs and remote healing could also soon be within reach. Taken to an extreme the technology will support the “no drivers needed” campaign. Google’s driverless car program has already completed 300,000 miles – it has been reported that the only accident was apparently someone running into the back of the driverless car. I would like to have seen the look of surprise on the offending driver’s face.
Getting into detail, the discussion on “how” to connect the car boils down to 2-3 options. There is already a lot of highways infrastructure that could be adapted to carry vehicle and driver information directly into vehicles. This mostly fiber-optic infrastructure is widely underused today in many countries. The solution could remove the need for overhead information gantries, as information would be on direct display in-vehicle, and would not be dependent on availability of mobile telephone networks. However the big drawback is that it is only available on main routes. So for now, the real options are (1) use your personal smart phone as a connectivity gateway, or (2) have a built in SIM card and modem in the vehicle. Both rely on existing 3G/4G networks and good signal coverage, which is also a limiting factor in many areas globally. By using your own smart phone, the benefits of a familiar user-interface, personal contact list, favourite apps, music and multimedia lists as well as a single consolidated data plan with your mobile service provider are significant. By using technology from organizations such as the Car Connectivity Consortium (Mirrorlink), the smart phone user interface can be replicated today on the built-in Infotainment head unit, which would have a bare-minimum of functionality embedded. This technology is already in production on over 15 car models, and is covered by GENIVI Linux offerings from organizations such as Mentor Graphics. Another significant advantage of this “remote phone” approach, is that updates can be frequent and trouble free, with up-to-date maps, navigation aids, and phone apps being available in-vehicle. BYOD (Bring Your Own Device) as it is becoming known, offers no guarantees on security, reliability, compatibility and the guest smart phone will need to link up with the Infotainment Head-Unit, via USB or Bluetooth today, or very soon via wireless link. This approach also presents a possible attack-surface for hackers, potentially putting the entire car electronic control systems at risk. Option 2 is to go to more trouble and cost as a manufacturer, and essentially embed a complete smart phone in the vehicle. This can be a lot neater visually as a solution, and can be much better controlled by the vehicle manufacturer for quality and guaranteed functionality. It could also have better signal strength, making use of external and existing vehicle antennae. The challenge though is in keeping the system updated both from a hardware and software perspective, and keeping the costs low. Looking at the rate of mobile phone development over the last 10 years, would anyone now want a 10-year-old phone permanently built into their vehicle?
Arguments can be made for both approaches, and I believe we can expect to see vehicles with both technologies on offer for a number of years.