Efficient management of mass data is an important aspect of autonomous driving

Over the years, the automotive industry has been dominated by technology and electronics. From the expansion of the Driver Assistance System (ADAS) to the dream of an autonomous vehicle, there is only one step that equipment manufacturers and manufacturers need to take. A revolution that generates and will generate huge amounts of data that must be recorded, transported and stored to create speculative software.

For several years now, the entire autonomous ecosystem, from manufacturers to garages, equipment manufacturers and dealerships, has focused on big data. Why? Simply because the technology makes it possible to create all kinds of data, to monitor the use and wear condition of vehicles, to manage their maintenance, to correct irregularities in motor driver’s driving.

The beginning of an automobile revolution that will, in the near future, give rise to the emergence of autonomous driving. From the use of ADAS-type systems (Level 1) to the possibility of a complete journey without any human shelter (Level 5) is defined in 5 levels, Autonomous driving was previously allowed, the most advanced in this regard, including France, the highest level 2. In, that is to say, the adaptive cruise control lane support and the steering wheel combined with the driver’s hand. From July 14, this human control will no longer be required, mainly on motorways.

However, motorists on all public roads still have a long way to go before they can benefit from Level 4 or 5 autonomous vehicles.

Data transport and storage: a big challenge

Because the biggest challenge in the development of autonomous vehicles is data collection and management. To simplify, the artificial intelligence of such a vehicle must be able to recognize the situation in real time and make the best decision of the result. However, in order to “train” AI in the best possible way and enable it to respond to all imaginable situations, fine data management is essential.

But here it is: right today, a research vehicle for Level 3 autonomous driving equipped with dozens of sensors (radar, leader, etc.) and cameras that all generate data in a variety of formats. However, as Seagate’s “Mass Data on the Go” report shows, this type of car generates approximately 150 TB of data per day – equivalent to about 38,400 two-hour HD movies – which must be transferred to an ad hoc artificial intelligence learning center.

According to Seagate, using a standard Gigabit connection, it will take up to 150 days for a fleet of 10 vehicles to make the transfer. Imagine how much data you would need to receive before reaching Autonomous Driving Level 5: The challenge is huge.

To address this, Seagate has created mass data transfer and management solutions, including live mobile arrays, which were unveiled at the Future Mobility Campus Ireland showcase event late last year in a prototype Jaguar I-PACE autonomous vehicle.

In concrete terms, this next-generation 92 TB storage solution, installed only on the trunks of these vehicles, quickly transfers the huge amount of data generated (5 GB per second, or 150 times more than a home fiber optic link of 200 MB), and securely. (AES 256-bit encryption), will be analyzed and processed by artificial intelligence and machine learning in on-premises data centers and / or in cloud-based dataleks.

Physical transport of data is often much faster than trying to send large amounts of data over existing networks. Physical data transfer solutions are also ideal options for data protection. Thanks to “data-at-rest” encryption, they provide a high level of security at the hardware level. It is more efficient and energy efficient than encrypting data on the software side for transporting over a network.

This is enough to answer the subtle question of mass data management and transfer. There are no conditions for autonomous driving to go from dream to reality.

To learn more about mass data storage solutions for the automotive industry, click here

Leave a Comment