As we evolve from the Connected Vehicle available on the market today, to the Autonomous Vehicle available in the future, we need to evaluate how data will be processed and analyzed to allow the vehicle to function free from human intervention – while delivering the driver safely to the desired destination.
The Connected Vehicle already generates a tremendous amount of data. It’s hard to believe that an Autonomous Vehicle will generate up to 1 terabyte of data every second (Tom Coughlin, Coughlin Associates, Data storage consultant). If the data is to be leveraged, it must be used and enriched on board the vehicle before being transmitted. Even Dedicated Short Range Communication (DSRCs) such as VANETS, MANETS, can benefit from enriched data. Cellular and other data transfer methods are not the solutions by themselves as the data volumes will overwhelm even the 5G networks being designed today.
So how do we address the large data volumes and still meet the objectives of the Connected and Autonomous Vehicle? The data volumes are likely to increase as systems become more sophisticated. It’s unlikely that all the data needs to be sent back to a centralized system - hence the entry of edge/fog computing systems.
Enormous amounts of sensor data, critical local processing power, and an equally essential need to connect back to more advanced data analysis tools in the cloud, make Connected and Autonomous Vehicles a very good use case for advanced Edge Computing. By moving the intelligence closer to the source, it is possible to enrich the raw data, create metadata, and send the information relating to the vehicle actions and decisions to their appropriate destinations. This will have a significant impact on the vehicle as the data becomes information faster, and the car becomes information rich (not data rich).
We also will need speed, as an Autonomous Vehicle cannot wait for transmission of data to the cloud and back to perform a critical function. We would want the data to be processed as quickly as possible. By moving Machine Learning functions to the camera, the computer vision, recognition, and the signal to apply the brakes can all be processed on the device where the data is generated, not collected.
Thus, Connected and Autonomous Vehicles are more than edge devices or mobile data centers or distributed devices connected to the cloud. They are sophisticated computing ecosystems with multiple networks, layers of security, and an exponential growth of processors which leverages Edge Computing devices to manage the volumes of data being generated and critical decisions being made. By leveraging Edge Computing, moving computing functions and machine learning closer to the data generation center, Connected and Autonomous Vehicles will continue to evolve at an exponential rate with the ability to function in the Vehicle-to-everything (V2X) environment envisioned by many OEMs.
To learn more about the various scenarios where autonomous vehicles could leverage Edge computing, download our white paper on “Connected and Autonomous Vehicles – Edge Computing Devices or more?”