FiveAI is a UK technology start-up bringing together the best minds in AI, engineering and mobility to deliver a fully autonomous shared transport service for Europe’s cities. Operating out of six UK locations, the company already has prototype cars out and about in the UK, and is planning its first full service trial for commuters in London by early 2020.
Can I start by asking how the FiveAI technology was born, and how the company has grown up around it?
We started in late 2015, with us founders mulling over ideas for an interesting business to build deep technology. We fairly quickly fixated on the idea of applying artificial intelligence (AI) to a particularly large problem. We looked at different options and the most interesting was applying AI to transportation. This seemed to us possibly the first real place where we’ll see the mass roll-out of AI to solve a real problem, and make a real difference to people’s lives.
So we decided to build the business in the autonomous vehicle space, even though the vast majority of the effort that had taken place to that point was in the US and, to a lesser extent, in China. We’re here to build a serious player—in Europe, for Europe. Our cities and roads are so different here, a dedicated solution is a must.
We took our idea to the venture capital community, and managed to raise a seed funding round in July 2016. From that point we started to scale FiveAI as a business. We went from just six of us to 18 or 20 in our first year. That put us in a strong position to raise our Series A round, in which we secured 18 million. In turn, that’s enabled us to scale from 18 or 20 people to around 130 today. Now, we’re working to build a more deeply developed architecture, and are creating a vehicle that will work on public roads in the near future.
So that’s how you got off the ground. What impacts will people eventually see from your technologies?
Autonomous vehicles will transform Europe’s cities, bringing huge and wide-ranging benefits. Consider the status quo—many people still get around by driving several cubic meters of metal, which typically comprises four or five seats of which they often occupy just one. They park this car near where they live and near where they work—using up valuable public space. And, on top of that, they spend a considerable amount of time in traffic that typically runs at around 10 miles an hour. People across Europe spend 88 billion hours a year in traffic, and roughly 930 billion dollars on buying and operating cars. The numbers are absolutely gargantuan. Even in London, 64% of people own a car, and most of those are using them to commute to work. All of this takes its toll on people’s day-to-day, and on the environment. Clearly, something’s not working.
That picture will change as three technologies come together. The first? Ride-sharing. Seat utilisation is at roughly 1% today, and if we can move that up to 25% we’ll end up with a massive payback for our cities. Think less traffic, and freer flowing roads. The second technology? Low-emissions electric vehicles. Our air will be cleaner, benefiting people and the planet. And third? Full autonomy. This will make both ride-sharing and electric vehicles possible, like never before. Without full autonomy it’s very hard to get ride-sharing into a good economic place, because the cost of the driver is just too high. To unlock the benefits, technology needs to be in the driving seat. Also, if you think of low-emission electric vehicles 10 years on, we’re going to have to find a way in which those cars can trundle off to a charging centre. Again, full autonomy is the solution to that.
So we’ll need autonomous vehicles—but how difficult is it to develop them?
Autonomy is key but, perhaps unsurprisingly, it’s a complex problem. FiveAI is building a full software stack. First up, we specify the sensors, the compute and other crucial subsystems that are to be integrated into vehicles—which, other than that for the near term, are very similar to today’s passenger vehicles. We then take the outputs from all those sensors, and process the information to work out where everything is in three-dimensional space, how it’s moving relative to the vehicle, what it is, and what it’s going to do next. Then we plan our action and control the vehicle.
This technology is very geo-specific, and must be launched on a route-by-route, block-by-block, city-by-city basis, and always bottom-up, never top-down. We’re decades away from a driverless vehicle that can drive anywhere on the planet. It starts with specific routes across a specific city, serving specific neighbourhoods. Then, piece by piece, it will build a network effect across each city, to the point when eventually people can give up their car. In this sector, it turns out growth is very linear for a period of time. Then, when you reach that network effect, the growth goes vertical.
What’s the commercial potential for your business when that happens?
Europeans are spending more than 900 billion dollars on owning and operating personal vehicles today, and our assessment is that in 10 years’ time 50 billion of that could be delivered autonomously. It could be three times that, up to 150 billion dollars, or it could be half that. But it’s a fair estimate that between 5% and 10% of that market will be delivered fully autonomously across Europe in 10 years’ time.
Our objective is to build the biggest company in Europe doing this, and to gain somewhere between 20% and 50% of market share in Europe. So, we’re architects of a company that has the capability to draw in tens of billions of revenue progressively over the next 10 years, all while improving the lives of people living in Europe’s cities.
What’s the state of play now with the business and the technology?
We’re already testing our system on public roads. We have eight fully-equipped vehicles, with many sensors and a huge amount of compute—trillions of arithmetic operations every second. We spent nearly $250,000 dollars on each car and they’ve been public road testing since June 2018. So we have a functioning end-to-end spec that takes sensor reports, systematically determines what they are, and reproduces them in how to control the car.
At the beginning of 2020, we’ll trial our service with a selected group of citizens, inviting them to experience our vehicles and share their ideas and feedback. By mid-2022, we’ll launch our first fully autonomous service in London, giving everyone access to safe, affordable, green travel.
As you look ahead for the next three years, what do you see?
Obviously, our team is going to get bigger. We’re 130 today, and we’ll most likely double that by end of next year. We need to continue growing from there, too.
You’ll see a new vehicle platform, as well. Our current platform is a Ford Fusion. We’ll be replacing this with a larger vehicle, which will be fitted out to be closer in size, shape, function and experience to a consumer passenger service. We’ll employ user-first design, to create something that feels at once democratic and extremely comfortable and innovative.
Crucially, we’ve got to step into Europe’s cities and position ourselves as the leading service. We’ll be building amazing, high-realism simulations of these cities and will use these to ensure we can provide a service on the best routes, in all weather conditions, at every time of day, and in every lighting condition. We need to show citizens that our system meets these extensive safety criteria, in a way no competitor is going to be able to do.
On a personal level, you must just be incredibly excited by what you’re involved in?
Yes, it’s certainly exciting. And daunting, because the numbers are huge. This can’t be done on a small amount of capital. Our ability to do it is highly coupled to our ability to raise investment dollars. If we raise a relatively small amount of investment dollars, we grow a relatively small business. If we raise a large amount, we get a chance to build a huge business.
So far, this technology has really been a US phenomenon. The total investment in this space in the US is about 10.5 billion dollars—it’s massive. China has seen big investment, too. Until now, Europe hasn’t benefited. But we’re here to change that.
FiveAI has the courage to take the lead in Europe. To take on this mission and build a seriously significant company, we of course need the right tech, business model and support of citizens and transport authorities. But sheer will is also vital. The easy way out is to explain why this can’t be done in Europe. The braver, bolder route is to know that, while all the lessons show it’s deeply complicated, we’ve got what it takes. Europe is super-saturated with the talent to make it happen. And we have a massive trillion-dollar market here. These are exciting times.
Please get in touch if you would like to know how we work with and support privately owned businesses. My team and I can help with all your business queries including audit, tax, cyber, data and advisory including transactions and corporate finance. Colin Bates, PwC Partner and regional private business leader.