Looking ahead, one thing is clear: Rapidly evolving technology is critical to the future of fully autonomous driving. Recently, I wrote about what goes on under the hood of the latest automated vehicles. Others in my field have covered the incredible value automated driving will have on our society — on a global scale, it could save more than USD $5.6 trillion per year by preventing accidents, boosting productivity and reducing fuel costs.
But this change won’t happen on its own. To turn a vision into reality, our transportation ecosystem will need new levels of collaboration and a comprehensive set of technologies that span three critical areas: the car, connectivity and the cloud
Car to Cloud: Compute Is the New Horsepower
To start with, cars will need incredibly high computing power so they can quickly react to changes on the road. They’ll need to recognize objects, determine how fast and in which direction they’re moving and make split-second decisions about how to navigate around them. This requires a level of in-vehicle computing we haven’t seen before.
To give system designers exactly the level of compute they need, Intel delivers a portfolio that’s incredibly scalable, ranging from power-optimized Intel Atom to high-performance Intel Xeon processors. Rather than pursuing a single microarchitecture to handle everything, we’re designing products that will enable real-time decisions to be made across a variety of independent computing elements where the workload could run. With multiple domains of overlapping compute and sensor processing, workloads can be distributed with greater safety and security. Furthermore, our flexible architecture maximizes hardware and software reuse, so automakers and tier 1 suppliers can pursue countless design iterations that differentiate their brands and accommodate a wider range of market needs.
But even if we have the technology to build fully autonomous vehicles, consumers will not embrace them until self-driving cars earn their trust. An automated vehicle’s human-machine interface (HMI) is how the car and its passengers communicate with each other. Intel is making significant investments in primary and collaborative research to reveal insights on how automotive HMIs can facilitate and enhance trust interactions. Our research has led to key findings about the interactions that build trust. For example, cars should share what they “see” with passengers and give them multiple ways to provide instruction or request changes
5G: Communication in Milliseconds
As mobile data traffic surges, connected vehicles will be among the billions of devices competing for network bandwidth. To confidently support vehicle-to-everything (V2X) communications and other new experiences, transportation providers will need increasingly faster data transfer speeds and response times — not just in seconds, but in milliseconds.
The next generation of mobile networking, 5G, is expected to deliver one-millisecond, ultra-low latency with 99.999 percent reliability at 10 Gbps speeds. It will allow automakers to target several exciting new use cases, such as high-definition (HD) map downloads in real time, HD content for in-vehicle infotainment and over-the-air firmware and software updates. In addition, data from the vehicle’s sensors will be uploaded to the cloud to inform machine learning models.
The industry is already beginning to pave the path to 5G. At CES in January, Intel announced the world’s first 5G-ready development platform for autonomous driving as part of our new Intel GO portfolio. AT&T recently announced two test markets for its 5G network. Other global telecom leaders, including Ericsson, KT, LG Electronics, Nokia, SK Telecom and Verizon, are working on integrated 5G solutions that will provide advanced connectivity and transform the network.
A Powerhouse for a New Era of Data
Finally, transportation providers will need substantial cloud capabilities powered by an advanced data center. With each automated vehicle expected to generate about 4,000 GB of data per day, data centers will need to support unprecedented amounts of information. This data will allow the automotive ecosystem to act on new business opportunities, such as transportation-as-a-service (TaaS).
Even more important is the role data centers will play in artificial intelligence (AI) simulation and ongoing training. The work being done in this area will lead to driving models that will make vehicles truly automated. Because it is so memory intensive, AI simulation and training require scalable, high-performance and power-efficient infrastructure.
The Road Ahead
As the car moves to the center of the autonomous world, transportation providers must swiftly respond to — and grow with — market demands. Intel is leveraging our expertise in IoT to deliver the most flexible, scalable and secure solutions for automated driving. Our new Intel GO portfolio is accelerating the time it takes OEMs to bring automated driving solutions to market.
By collaborating with some of the world’s leading automotive brands, we are building solutions unlike anything seen before and helping transportation providers map a path forward. From vehicle dynamics to semiconductor physics, car to cloud, we are preparing the industry for the amazing future of transportation and truly automated vehicles.
To learn more about the road ahead for fully automated vehicles, visit intel.com/automotive. For more on Intel IoT developments, subscribe to our RSS feed for email notifications of blog updates, or visit intel.com/IoT, LinkedIn, Facebook and Twitter.
– Jack Weast, Principal Engineer & Chief Architect of Autonomous Driving Solutions, Intel
JOIN THE CONVERSATION
Share your thoughts and questions in the comment section below. To get the latest news from PCM, follow @PCM on Twitter, join us on Facebook, or connect with us on LinkedIn. To get the latest news sent straight to your inbox, join our newsletter.