Intel’s Patti Robb: ‘Autonomous vehicles will be data centres on wheels’

25 Oct 2017

The autonomous vehicles of tomorrow will be data centres on wheels that will communicate wirelessly with other vehicles. Image: Intel/Mobileye

The leader of Intel’s Silicon Valley Autonomous Vehicle Lab says that data and 5G will fuel the cars of the future.

It isn’t just the combination of silicon, 5G, data and sensors that has Patti Robb, the chief strategist in charge of Intel’s Silicon Valley Autonomous Vehicle Lab, fired up about the future of transport.

For her, it is personal.

“My brother passed away after sustaining injuries in an auto accident,” she said, explaining that her motivation is as much about safety and saving lives as it is about the march of technology.

‘Really, there has never been a more transformational time to be working in the autotech industry, and it will really transform economic opportunities for passengers, too’
– PATTI ROBB

It is also about freedom. “My uncle was in a wheelchair and struggled to get around and I watched the joy on his face when he got his first car and was able to have some freedom. The power of this technology to provide mobility to the disabled is also why I am super-passionate about this space.”

Intel’s Silicon Valley Autonomous Vehicle Lab works with tier-one suppliers, carmakers, universities and start-ups to accelerate delivery of autonomous driving solutions.

Revving up for autonomous driving

Intel’s Patti Robb: ‘Tomorrow’s vehicles will be data centres on wheels’

Patti Robb is a visionary technologist and the chief strategist behind the Intel Silicon Valley Autonomous Vehicle Lab. Image: Intel

Robb said that the autonomous vehicle revolution is revving up and that by 2010, more use cases will emerge in cities, campuses and urban areas.

“Widespread adoption is going to take a little bit of time but right now, the biggest factors getting the roll-out moving are people getting exposure and experience with the technology, trusting the technology and also policy.

“We are doing a number of things in that area; we are working just like we did in the PC area, where we rolled out countless platforms that propelled the PC industry. We are working with standards bodies to put standards in place and we are also working with partners like BMW who are inducting Intel Mobileye into their vehicles.

Robb is referring to the massive $15.3bn acquisition of Israel-based Mobileye, a leader in autonomous driving technology, which was concluded in August.

Crucially, says Robb, vehicles of the future will not only be transmitters of data via 5G to other vehicles with vital real-time information, but also repositories of data.

“There are three components to this: car, connectivity and cloud. The car of tomorrow, the autonomous vehicle, is really going to be a data centre on wheels. It is pretty fascinating to think about, isn’t it?

“When you think about it, the data centre is going to be collecting and analysing terabytes of data. There are hundreds of sensors on the car and, in the lab where I work, one of the things we are doing is we are looking at different types of sensors – camera, radar, sonar, GPS, LiDAR and more.

“Why do we need all of those sensors? When you think about it, you need to be able to see a 360-degree view of what’s around your car, and think of your sensors as your eyes and ears and the data centre processor as the brain of the car, and you need all of those sensors to see around you. If you are driving to work, there are so many things to come into contact with … and you need to be able to operate whether it’s sunny, rainy, when there’s snow and ice – and each of those types of sensors has their place. And, if you think about it, there are lots of different types of things – some are close up, like bikes and pedestrians, that your car needs to see even before you do.”

The $7trn passenger economy

Crucially, Robb believes that the notion of car ownership could change as the sharing economy is embraced by car manufacturers and not just unicorns such as Lyft or Uber.

“We are doing a lot of work around individual passenger economies and we are seeing differences in generational attitudes.

“Millennials are very open to shared usage but still, older generations want to own their own car and it is associated with personal freedom.”

Robb pointed to a recent study by Intel and Strategy Analytics, which valued the economic opportunity of the passenger economy as one that will grow from $800m to $7trn as autonomous vehicles become mainstream and new services and apps emerge to support the idle time when drivers become passengers.

“If you think about it, if you are driving in an autonomous car, you are going to have so much more time to do things with that time – watch movies, have a personalised experience around what you would like to buy or interact with. A whole app economy will emerge to envision all these different uses.”

Robb said that despite the advances in technology, the spread of autonomous vehicles will be determined by policy and, crucially, safety.

“The first use cases will be city by city. We may see dedicated lanes, geo-fenced areas … because, at first, you’ll have a mix of autonomous cars and cars of today.

“In my role here at the lab, we are setting up one of the first innovation centres in the industry to have abilities like our own 5G cell site here, data centre in the cars, automation through test drives; and we look at 5G use cases for fast download and connectivity to get data from the car to the cloud with as least latency as possible.

“As well as use cases, some of the other things we are learning about are high-definition mapping and updating data – like if your car sees a pothole, it will send that information back to the cloud so other cars can learn that there is a pothole ahead and avoid a collision or accident.”

When it comes to data, regions will need specific algorithms. “Europe and roundabouts and more left-hand turns,” she laughed, recalling a recent trip to Ireland.

“The autonomous vehicle landscape will require US-specific and Europe-specific algorithms. With high-speed data offloading from the car to the cloud, we need to determine what data should stay in the car, what data should go over the network to the cloud, so we can do machine learning on that, and also be able to do deep learning and build models and take that data and pass it back to the car, so the car can learn from it and handle different driving scenarios.

“Our test vehicles allow us to understand the different compute environments, and, by building and testing the technology first-hand, we are going to be better positioned to help our customers to deliver those autonomous solutions.”

5G and securing the roads of tomorrow

No doubt the spectre of hackers infiltrating vehicles has fired up the popular imagination but, when it comes to the future of autonomous vehicles, it really is a life and death issue.

“Security is always top of mind for Intel, and our hardware security and software security solutions are being designed to protect the vehicle’s data, and we are continuing to work to improve the authenticity and integrity of data and how it is transmitted across the network.

“If you think about it, this isn’t a new problem. We’ve solved this problem for security in PCs, in smartphones, and we have a lot of people using our data centres today for very trusted applications like banking. We need to continue to work together as an industry, and technologies need to evolve to address threats and enable automakers to continuously improve security features on a vehicle in ways which they’ve never truly had the ability before.

“Our technology supports resilient, flexible security architecture that can take advantage of those over-the-air updates.”

Robb believes that safety will be the biggest contribution that autonomous driving technology will make to the world.

“There are three main contributions that autonomous driving will make: the largest one is safety, second is mobility for everybody and third is traffic reduction.

“The additional one will be the transformation of the economy; how the economy changes as a result of this and the new businesses that will spring up that don’t exist today.

“We’ve talked about how the passenger economy is spurring a potential $7trn economy – there’s going to be so many new businesses that will crop up, we probably cannot predict what they will be.

“Will there be a dining car that will come to me instead of me going to the restaurant? Or my personal AI assistant will know where I want to go, if I want to make a quick stop at a Starbucks, and the coffee will be waiting for me? It’s quite fantastic the kind of businesses that could spring up around it, enabled by all the different technologies it works across, such as robotics, machine learning, working with new types of mechanical actuation systems, 5G and [the] great potential we have there.

“With 5G, we unveiled our first 5G platform at Mobile World Congress last year to demonstrate over-the-air interoperability with Ericsson, with different use cases, low latency, sharing mission-critical information, vehicle-to-vehicle information, HD maps that come with the car. There’s a tremendous amount of things that will happen over 5G as well.”

Robb concluded: “Really, there has never been a more transformational time to be working in the autotech industry, and it will really transform economic opportunities for passengers, too.”

John Kennedy is a journalist who served as editor of Silicon Republic for 17 years

editorial@siliconrepublic.com