Not long ago, innovation within the automotive industry was about building more powerful engines, then increasing efficiency, while improving aerodynamics, comfort, and the exterior design of vehicles. Today’s automotive industry is all about hyper connectivity and automation.
But, says Christophe Vaissade, sales director EMEA, Western Digital, while driverless cars are likely the first thing that springs to mind when you think about the car of the future, the future of the industry expands beyond driverless tech.
Connectivity is a major factor in the transformation of automobiles, enabling remote updates, predictive maintenance, wider safety and data security measures. And data collection and storage are at the heart of it.
While increased vehicle connectivity has improved the overall driving experience, the connected car gathers, processes and generates vast amounts of data. Last year it was estimated that the storage capacity requirement for autonomous vehicles would balloon to over 2TB in the next decade and, as technology continues to advance, that number could grow. With that in mind, original equipment manufacturers (OEMs) will be asking how they can support these rising data demands.
How will the architecture of autonomous cars develop?
Machine learning and artificial intelligence models are crucial to managing the data of autonomous vehicles and to improve features like object detection, mapping and decision-making. By improving the algorithm of the machine learning model, manufacturers will be able to enhance the user experience.
The architecture of autonomous cars is also becoming much more streamlined. Manufacturers are moving away from a spread-out network of microcontroller units (MCUs) for each individual application, and towards one central MCU processor with big compute capability to meet data demands at both the edge and in the cloud. Where that data goes depends on whether it’s needed for the immediate benefit of the driver. Information collected by motion sensors or the mapping function of the GPS can support improvements made to the ADAS system.
The role of 5G in the future of connected cars
While today’s 4G networks will continue to handle many applications, 5G could be a key enabler for connected and autonomous vehicles, as they will be able to almost instantaneously communicate with each other, buildings, and infrastructure (V2V, V2I, V2X).
Connectivity is key for autonomous vehicles and 5G will be able to offer faster connectivity and lower latency for drivers of the future. Faster speeds will mean much quicker response times for the data that the car collects, allowing the vehicle to react almost instantaneously to sudden changes in traffic or weather conditions. 5G will also enable digital services from within the car, thereby enhancing the passenger experience and increasing revenue potential as a result of such services.
Data security: Who holds the keys?
Without a doubt, autonomous vehicles need state-of-the-art cybersecurity. According to a recent study, which surveyed auto engineers and IT experts, 84% of respondents were concerned that car manufacturers are not keeping pace with the industry’s constantly increasing cybersecurity threats.
To protect consumer safety and privacy, connected cars must be secure at all levels — from the hardware and software inside, to the connections to the network and cloud. Data security can be achieved in many different ways; manufacturers should ensure that data of all forms are encrypted, so that access is limited to authorised parties only. It’s also critical to implement end-to-end security to protect every access point of the car, from micro sensors to 5G masts.
The importance of a back-up plan
A car’s central data storage system is mission-critical, so how can manufacturers ensure that operations are not compromised if this system fails? Replication of the data into a redundant data centre system is one way to prevent accidents if the main centre fails, but this can be incredibly expensive.
Some engineers are developing back-up systems for specific components of the car, such as the brakes, steering, sensors and computer chips that guide a self-driving vehicle. This way, the vehicle is equipped with a secondary system that will safely pull the car over in case of a catastrophic equipment failure, without the need to back up allof the data that is stored by the vehicle. Other car functions that aren’t as critical and don’t require a back-up, such as air conditioning and radio, can be left out to help cut costs while also providing a safety net in the event of systems failure.
As we move toward the autonomous vehicles of the future, data will be driving the evolution. By leveraging machine learning solutions to support the huge amounts of data these vehicles depend on, and by implementing robust security strategies to protect the car from the outside world, manufacturers will be designing cars that will eventually be safe enough for the newly paved digital roadways.
The author is Christophe Vaissade, sales director EMEA, Western Digital.