In our latest Founder Feature, we talk to Mark Nicholson, CEO and founder of Vivacity Labs
Mark co-founded Vivacity Labs with Yang Lu and Peter Mildon, who met while designing, manufacturing, and racing a road-legal solar-powered car. Subsequently, Mark worked at a top-tier strategy consultancy, supporting board-level decisions at a range of household brands, including Pizza Express and Hertz.
Now, Mark combines his technical and commercial experience with a passion for bringing artificial intelligence to the transport systems we all use every day.
Read Mark’s fascinating founder journey below.
Q: Could you tell us about your company and what you’re striving to achieve?
At Vivacity, we’re helping to modernize the way cities understand and optimise their road networks. This spans everything from quantifying undertaking behaviour at a junction to key insights into strategic planning; from an alert that a vehicle has broken down to perfecting traffic light timings with deep reinforcement learning.
To solve these issues, we use AI on sensors to gather data, and yet more AI to control key parts of the infrastructure. We’re now in 30 towns and cities across the UK, and expanding rapidly to give transport authorities the tools they need for a 21st-century transport network.
Q: Who are you and what is your story?
The story starts back in 2011, when I met my cofounders while designing, building and racing a solar powered car. It was a student-run project, done in our spare time alongside our degrees – in which we raised £500,000, built a road-legal experimental vehicle which was 50x more efficient than a normal car, and took 20 people to Australia to race it. While that project didn’t lead to the success we wanted, it gave us the entrepreneurial bug, and so after a couple of years in “normal” jobs we got itchy feet, which led to us founding Vivacity.
Without that solar car project, we simply wouldn’t be here today – not just because we wouldn’t have met, but because we wouldn’t have learnt the critical lessons needed to get a company off the ground. Learning those lessons wasn’t always easy (we left £10,000+ of solar panels in dust on the road after an 85kmph crash!*), but they were formative.
* The driver walked away from the accident, without significant injuries
Q: Tell us about some of the biggest obstacles you’ve had to overcome?
Road transport infrastructure is safety-critical – you can’t have two traffic lights in perpendicular directions going green. Introducing black-box AI systems into that ecosystem is hugely challenging; not only do we need to convince ourselves that a probabilistic system will never do anything unexpected, but we also need to make it sufficiently interpretable that it can be defended in court cases and in public scrutiny. Alongside the technical challenge, we also needed to persuade the industry that AI should be deployed for safety-critical purposes in the real world.
To get there, we started by building credibility on simpler work, deploying sensors to gather novel datasets for clients. Those relationships then let us debate with clients how this might work, before seeking the funding to do it properly. When it came to deploying something, we went through the full safety-critical process – doing everything in simulation, then in the lab,
long before deploying in the real world. It’s about finding that balance of a startup’s desire to “move fast and break things”, vs the reality of the industry we’re working in.
Q: What’s most exciting about your traction to date?
I think one of the most exciting moments was the first time that we deployed AI to actually control real traffic. All our work over 4 years on the sensors, and 18 months on traffic control, was finally crystallised into something journalists had hailed as the future of AI for over a decade – actually having AI managing traffic lights! We’ve continued to iterate the technology since then, demonstrating AI-superiority in controlling traffic, scaling to multiple junctions, and soon to be multiple cities. But that one moment stands out as a huge achievement.
Q: What is your plan to adapt if your industry is completely disrupted?
I think startups always have an advantage when it comes to disruption – because we remember what it was like to be inventing a company from scratch, so reinventing ourselves isn’t as hard as it is for a corporate. For us in particular, our underlying technology is really flexible; essentially we have a very capable set of computer vision algorithms, internet-of-things hardware and control systems, and optimization algorithms. Any ecosystem where you could gather data from cameras and use that to optimize an operation would be a great fit. We’ve done some blue-sky brainstorming, which has come up with a huge range of wacky ideas, from space to agriculture – and I’m sure our developers would love the challenge of a 90-degree pivot!
Q: What are the future implications of the technology you are developing & how would you deal with that?
Whenever you deal with AI, you always need to be aware of the misuse of your technology – from virtual world to deepfakes, improved insurance optimization to systematized discrimination. For us, I think the black mirror technology is China’s panopticon. We set out on our mission to provide a citizen-first, privacy-by-design “Smart City”, but China has demonstrated how a lot of this technology can be abused to create a surveillance state, with horrific ramifications for the Uighur population.
We believe that the only way to hold true to that mission is to make sure you are accountable. Any project which treads close to the grey line between right and wrong is put to a debate and vote across the company – and we don’t proceed unless our employees are comfortable. This mission, and approach, has led to us hiring a team which is being really passionate about that line, which is in turn self-enforcing on our overall direction.
Q: Do you see this as UK centric or will you conquer the world by going global?
Globalisation is always a tricky question, which for me comes down to the complexity of the product, and the synergies between different countries. For us, releasing deep tech products with huge amounts of research poured into optimizing traffic, where each country has fundamentally the same problems with congestion, air quality, and prioritizing sustainability (eg cyclists, buses), the product just has to go global. Staying in the UK isn’t really an option.
Looking at the historical precedent here, the UK invented one of the world’s leading traffic control algorithms in the last generation back in the 1980s, which is now installed in many hundreds of cities worldwide, but that is now the best part of 40 years old. We want to bring the UK back to the forefront of innovation in transport infrastructure, leveraging the UK’s amazing research base and globally respected institutions to support that.
Q: What’s one piece of advice you’d give to budding innovators taking the same journey?
My biggest learning has been to trust my gut instinct and act on it – particularly when it comes to getting the right people in your team. We’ve hired people with no credentials but who had a sparkle in their eyes, who have turned out to be superstars. Similarly, we’ve hired people who caused minor issues that raised eyebrows in their first few months, which turned into major rows and big problems over time. You need to be reasonable, and be humane, but often the best thing for you and the person in question is to part ways earlier, even if it feels painful to lose that resource and look for a new person at the time.
Find out more about Mark Nicholson and Vivacity Labs here: vivacitylabs.com