Self-driving cars will generate huge amounts of data (potentially in the form of terabytes of data in an hour). This can hamper response times when the vehicle needs to respond to an object, it is also one factor why self-driving cars could turn out to be very expensive. This is due to the expense of bandwidth and also, RT Insights assesses, with getting latency within reasonable limits. This has fueled an interest in edge computing as an alternative.
READ MORE: How edge computing can boost business efficiency
With autonomous cars, some data will be transferred to a cloud and some data will need to be assessed locally. With this distinction, data that is critical to safely navigate the car needs to be analyzed locally by the vehicle itself, and for this to happen edge computing is essential. Edge computing aids autonomous vehicles to achieve situational awareness. Furthermore, the use of edge computing boosts reliability and the recall of data.
Charging points for data transfer
One example of how edge computing is being used in this way with autonomous vehicles is with the partnership between edge data center operator EdgeConneX and Renovo, which develops a software platform for autonomous vehicles. The two companies have developed a means to handle the bandwidth and latency problems.
EdgeConneX is putting computing power into places that cars will regularly visit and stay in for long-enough stretches of time – the charging points. As well as providing power to cars, the charge points are also being configured to enable data transfer. This means the ethernet connection transmits data to local storage far faster compared with an attempt to upload data to a cloud service.
Vehicle to vehicle communication
Autonomous vehicles collect considerable volumes of information via sensors. To aid fleets some of this information can be shared via vehicle to vehicle communication systems. Transmitting key data such as weather changes and road conditions, enables other vehicles to understand potential hazards such as detours, debris, and accidents. If such information is analyzed and assessed enough then other vehicles can make the appropriate adjustments.
There will also be a need to assess some of this data in more detail, to help fleet managers to assess conditions and to enable computers to learn. For this to happen, vehicles will need to have access to edge data centers to offload critical data for later analysis.
Smart cars in smart cities
Many urban planners see autonomous cars as central to designing the smart city of the future, and edge computing is pivotal to realizing this vision. An example is with improved surveillance, such as the use of a vehicle tag recognition app that sits on an edge gateway and which would enable police to scan parking lot cameras in order to pinpoint a certain car.
