Imagination Technologies is a company with a vision to help create great, innovative products that will change the world. We are always thinking ahead to see how we can best deliver that future – a future that’s bright, bold and empowering. At the core of this vision are our employees who are key in bringing this future to fruition and have an interesting take on the world.

We, therefore, present a series of interviews with some of our key thinkers, where we’ll learn a little about their background, what makes them tick and then find out what they see coming down the road. Following on from our first in the series that focussed on the smart home and the second that looked primarily at vision, for our third interview, we look at the world of car. The automotive landscape is evolving at a great pace and many of Imagination’s technologies will play a role in shaping this future. We talk to Bryce Johnstone, Director of Ecosystems, Segment Marketing for Automotive, to give us an overview of where we’re headed.

Bryce Johnstone, Director of Ecosystems, Segment Marketing for Automotive

Bryce, tell us a little about your background, how you got into technology and your journey to Imagination.

I went to Edinburgh University and ended up doing one of the very first microelectronics degrees in the UK, so that led me into a career in semiconductors. I led a team at STC Semiconductors working on devices that were in every British Telecom handset; there was a point where if you bought a phone it had my team’s tech inside.

I then joined Texas Instruments where I held a number of roles supporting a wide range of microprocessors: JTAG, DSP, x86 and SPARC. I helped Sun earn its biggest SuperSPARC design win in Europe and later I went into the fledgeling wireless terminal business unit, which grew from early beginnings to a $5.5 billion-a-year business unit. From there I went on to working with OMAP and then moved over to Imagination.

What’s your role here at Imagination?

My role here is to talk to the whole automotive ecosystem and find out the cool stuff that they’re doing. This helps us build up a radar map of what’s coming up so we can provide the technology needed and make sure that we align with the future requirements of Tier 1 automotive suppliers and car OEMs.  That’s our challenge – to work out what’s required three to five years in advance. Our job is to do some fundamentally smart stuff – or else the automotive ecosystem doesn’t happen.

What are the main challenges facing the automotive industry?

There are three areas it is working to improve: congestion, pollution and safety. First, congestion: there are more people moving into urban centres and more cars, and it’s going to get worse.

Shanghai is a classic case. It has a population of about 30 million and three ring roads and they are all pretty much jammed solid every day, so a short journey in Shanghai of maybe five miles will take you about an hour.

Congestion
Congestion is one of the issues that the automotive industry is looking to tackle.

Then’s there’s pollution. The sheer number of cars in many cities globally, particularly in Asia, has led to serious health issues for the population.  This is driving a range of measures such as using ‘lean burn’ technology to increase efficiently and reduce CO2 emissions and for the full electrification of cars to increase efficiency and reduce pollution by an order of magnitude.

Third, there’s safety. The main thing about automated driving is that it’s trying to stop 1.2 million people a year being killed in road accidents – most of them in the sub-continent, where they have an appalling track record – almost half of the road deaths globally.  Today, over 94 percent of accidents are driver error, while three percent of the world’s gross domestic product is spent cleaning up after them. So there is a problem and it needs fixing.  First, we need to take the driver out of the equation with trusted technology that is failsafe.

traffic India
Almost half of road deaths globally are in the sub-continent

What work is being done to help solve those challenges?

If you’re a government facing austerity then you’ll want to invest in things that will reduce that bill. The response is government initiatives and funding. In the UK, for example, we have UK Innovate and the Centre for Connected and Autonomous Vehicles (CCAV). These are intended to build up British expertise in automated driving, from algorithms and Convolutional Neural Networks (CNN) development to the building of testing environments and hardware.

This CNN development is critical. With autonomous driving, the further up the level of autonomous car control you go you have to multiply the performance required for each step by a factor of ten. So actually predicting pedestrian behaviour needs a 1000 times increase in performance and a much deeper set of algorithms.

CNNs
Advanced Convolutional Neural Networks (CNNs) are seen as critical to the development of autonomous driving Technical University of Munich

Attribution: Technical University of Munich

We already know what is likely to happen; this information is available. Every accident in the UK is logged along with details such as road conditions, the speed and direction of travel, if the driver indicated etc. You run this dataset past algorithms and a set of coefficients come out, which get ever more accurate; it’s quite mind-boggling.

The nirvana is that all cars will be fully autonomous and there can be no accidents. However, analysts are saying that it’s not until 2035 that we’re going to see any significant amount of such vehicles. Managing the crossover between cars with no ADAS functions and 100 percent full autonomy is going to be a huge challenge.

How will these developments help improve and change the automotive world by 2045?

Autonomous driving will also help reduce infrastructure costs. In the UK, the government is spending £1.3 billion to build up our road network. That buys you just a hundred miles of motorway. Autonomous cars with Vehicle-to-Vehicle communication (V2V) can talk to each other and respond within nanoseconds. So if one a car dabs its brakes then all the cars do, so you can move the cars closer together giving you a higher vehicle density on the road. It means you don’t necessarily have to build additional lanes.

Congestion needs to be handled differently within cities and eventually, we will see the introduction of Cars as a Service. London is a nightmare for congestion: I know people who have moved out as they can’t park their car within half a mile of their flat. Again, you can sort that out.

Big mega-cities with fully integrated transport systems could declare that if you want to drive a car within the city limits the car must offer Level 4 or 5 autonomy – nothing below can enter. The car could drop the person off and then go and park itself – and at the end of the day, it’s right outside waiting for you.

2030
Fully autonomous driving is not predicted to be commonplace until a 2030 timeframe

Attribution: Freescale/NXP

What’s more, 30 percent of road trips in cities are simply to find parking spaces. Using surround view cameras, cars could identify parking spaces, and that information can be sent up to the cloud and centrally cloud-sourced. So if you want to park somewhere there’s a database that tells you where you can go. We are already seeing multiple apps for this.

If you have a fully automated system and you’re living in the city then why bother with owning a car? Today, I pay insurance, I pay maintenance, road tax, etc., and I use it for an hour a day, so 23 hours a day my car is costing me money.

There are schemes, such as the Cadillac program in New York, where for a fixed monthly fee you get access to a number of models of car that are delivered to your location as and when you need. In a truly smart city you could call up an autonomous car and it would take you somewhere, and you only own it for that period of time – much more efficient. It would massively reduce the cost-of-ownership.

So car ownership will change and the car manufacturers are beginning to understand that. For older people owning a car is aspirational – but millennials they don’t care, it’s not a status symbol.

The car companies are beginning to get their heads round this. This is why, Ford, Volvo have a deal with Uber and GM have a deal with Lyft. Ford are going to delivering a Level 4 car for Uber by 2021, so the Uber driver becomes like the pilot of a plane – 90 percent of the drive is autonomous but the driver just has to do something at the beginning and at the end of the journey.

But Uber wants to get to the point where it has no taxi drivers, as 70 percent of what it pays out is to the driver, so get rid of that and profits go through the roof. So there will be a knock-on effect on jobs. In 50 years there will be no taxi drivers. And there’s currently a worldwide shortage of truck drivers; autonomy solves that problem too.

Is this all just a matter of time? What are the challenges in making these improvements happen?

A key issue is legislation. What will hold back autonomous driving are the liability lawyers in the U.S. For example, issues such as, what happens when a non-ADAS vehicle smashes into an autonomous vehicle? Is it always the fault of the non-autonomous vehicle? Is there a circumstance where the autonomous vehicle could be in the wrong? The way I look at it is that autonomous vehicles can’t cause a crash, but I imagine there could be circumstances where that might not be the case. The consequences of these scenarios are issues that will need to be worked through.

What is Imagination’s role? Where do we sit in the automotive value chain?

We’re a technology enabler. We sit at the base of the multiple layers of technology that play in this area. We are the foundation of a multi-trillion industry.

connected cars landscape
Imagination’s IP underpins the many layers of the connected automotive eco-system

We are pivotal in enabling it to exist. Our MIPS microprocessors will run the code that tells you what is happening in the car, while our Ensigma products will be sending that up wirelessly to a service provider. And when the car controls itself your dashboard or your windscreen can be turned into a display. You can surround yourself in a virtual world if you want to; that’s where our PowerVR GPUs come in.

Then there’s vision processing. While older cars on the road have no cameras a high-end car will have up to 20, so, our PowerVR GPU Compute and ISP technology will come into play.

Surround view Mercedes F-015
The Mercedes F-015 concept car has displays all around the car

We also need to think about hardware-backed security to prevent hacking. Now, for the mindset of the hacker, they don’t just want to take control of a single car as that’s not very interesting, but if they can say, get into every single BMW 5 Series and drive them all off the road, then that’s what they will try to do. Our OmniShield technology, (recently nominated for a National Technology award for ‘Security Innovation of the Year’) provides a hardware foundation for security-by-separation that we believe will be crucial in autonomous vehicles.

I think that the automotive industry has been much maligned in recent years but now it is fundamentally working towards solving big issues; congestion, pollution and saving lives. It’s going to be a massive force for good and I’m very glad to be a part of that at Imagination.

 

Comments

  • Richard Woollaston

    I feel Imagination may have missed the boat. Last November, Nvidia reported significant revenue growth fuelled by demand for its DRIVE PX2 platform. Here is IMG taking a long distance view of the nature of the market. It would have been nice, particularly in view of the recent news, for IMG to be reporting robust revenue growth in this area, rather than aspirational blue sky thinking. If IMG wants to deliver solutions then it needs to build and sell them. If it wants to sell IP then it needs to focus on selling the technology to those who already have done the blue sky thinking, rather than preaching to them.

    Sorry to be so curt, but one of the problems of the past was, in my view, IMG with its solutions focus trying to tell its customers what they should do rather than selling what they needed.