1. If you could direct £50m to drive a bold and focused scientific/technical step-change in human progress, what would you do?

I would fund 10 two-year projects and set 4 annual challenge prizes to develop, trial, and discover products and methods to deliver better public transport in large British cities (populations 1-5 million, total population 15 million) without large infrastructure investments. Imagine a version of DARPA’s grand challenge for autonomous vehicles but for getting 80 people on and off a bus in Wigan as quickly as possible while collecting their fares. Imagine a Kaggle-type prize for designing the economy-optimal set of profitable bus routes in Liverpool.

Five projects would extend their research for two more years. Two would consolidate the best ideas and sell their products and expertise to the large British cities and beyond.

Why this problem?

Among international comparators large British cities have the worst combination of public transport and roads. Their populations are isolated from the labour market by congestion and slow unreliable journeys.

Partly as a result, these cities are the least productive in the rich world. Their underperformance leaves the British economy the weakest in North Europe and North America.

Recent changes in British law frees these cities to fix their transport. Starting in mid-2024 Manchester will specify and run buses, set and collect fares, design routes and timetables, and force the creation and use of simple integrated ticketing systems. Others will follow. But after fifty years of these activities being banned, we lack the technologies and tacit knowledge needed to succeed.

We must relearn how to deliver more public transport journeys with fewer vehicles and drivers. We must relearn how to get people on and off vehicles quickly, how to design bus routes, place bus stops, and design ticketing systems that make travel easy and efficient. We must relearn how to collect data on people’s journeys so that we can continually improve.

This relearning will occur under substantial constraints. The large British cities have almost no power to raise taxes or issue bonds. They have no R&D budgets to fund research themselves and no power to compel local universities to work with them on their issues. Their new transport powers require a consent from UK government that can be unilaterally withdrawn.

In such unique circumstances we cannot learn much from abroad and there is a strong disincentive for university researchers to undertake early work that will be of only regional significance and thus score poorly in the REF. While the research and trial outputs are likely to be commercially viable in four years, there is not yet a market for them to attract private investment.

ARIA can invest here where others cannot.

But ARIA is also why this project risks failure.

The UK’s large cities will host the projects we fund and be early customers of the companies this project generates. They will be sceptical that an organisation with no presence, no board members, and no advisors in their places is committed to their success or able to build a core team that would usefully and collaboratively focus on their challenges. I share their scepticism.

2. On a scale of 1-10, if you were selected as a Programme Director, how likely would you be to build a programme around this specific vision vs. another topic?

I cannot imagine ARIA providing me with any advantage to work on any other topic. But you never know.

3. Who are three world experts you would love to talk to further explore this vision?

Past and present constraints on the provision of public transport in the large British cities are so unusual that there are few relevant world experts. Existing experts will tend to bring their previous experiences and apply them wrongly to our problems. This is why funding must be assigned to prize funds that attract new people with new technical solutions. The winners of these public challenges would immediately become a leading world expert in the field.

If I spoke to three experts I would speak to the following, focusing very narrowly on these topics.

• Jean Tirole. On how an unsubsidised integrated bus network might operate profitably.
• Jarrett Walker. On the challenges of working with politicians in Dublin on their bus network redesign.
• Sarah Boyd. On how Lothian buses run a low fare bus network with low public subsidy and why large British cities have failed to copy them.

4. What’s your primary area of scientific or technical expertise?

I have a degree in Physics and a PhD in Computational Biology. I started my own software and data business after my PhD and I have since founded, worked for, and worked with five startups; one acquired by Google, two VC funded and growing rapidly, and two continuing to grow organically.

The common theme of my work is the collection, high performance processing, and application of large datasets to economic and social use cases, often involving transport data, machine learning, geographic data, and economic analysis.

In the very small overlap of big data collection, high performance computing, machine learning, economic analysis, and transport data I am among the world’s best, with a record of enabling others to do even better than that.

5. How would you describe the depth of your expertise to someone else in your area?

I wrote software that tracks every bus in the West Midlands and I’ve been running it for seven years. I invented the technique that converts what the buses actually did into a timetable that we can compare with the published timetabled (which I led a team to convert into an analysable format). Using this technique, I showed that at peak time Birmingham’s effective population is less than half what we’d expect from its population density. I then invented the technique that suggests this may explain up to two thirds of Birmingham’s productivity shortfall versus comparable cities around the world. This work was featured by The Bank of England, in the media, at scientific conferences, and in scientific papers. My effective population techniques have now been applied more widely and improved upon by others who continue to inform UK and foreign policy. Most importantly they have improved bus services in Birmingham.

6. How would you describe the depth of your expertise to someone outside of your area?

I wrote software that tracks every bus in the West Midlands and I’ve been running it for seven years. I invented the technique that converts what the buses actually did into a timetable that we can compare with the published timetable. Using this technique, I showed that at peak time in Birmingham fewer than half of the city’s population can get to the city centre within 45 minutes. I then invented the technique that suggests this may explain up to two thirds of Birmingham’s weaker economy compared to similar cities around the world. This work was featured by The Bank of England, in the media, at scientific conferences, and in scientific papers. My effective population techniques have now been applied more widely and improved upon by others who continue to inform UK and foreign policy. Most importantly they have improved bus services in Birmingham.

7. Link to the highest impact, single piece of evidence to support this

Is all economics local? - speech by Andy Haldane | Bank of England

8. Share one example of when you’ve brought a self-generated scientific or technical vision to life.

I am the CTO and co-founder of The Data City. Over the past five years we have built a team of fifteen that continues to reinvent industrial classification and make it relevant to a modern economy of high value services and service-integrated manufacturing. Among many innovations, I have led a team that has invented a system for classifying what millions of companies do from their full website text, totalling well over 50GB of data, in less than five seconds. We currently provide this service via the web to up to 100 concurrent users with no performance penalty and with a proven capacity to expand to 1000 concurrent users within existing infrastructure.

We have faced huge challenges during our growth as we have improved the performance of our core classification technology by a factor 20,000. We have dealt with, worked around, and eventually fixed bugs that took over a year to properly diagnose. We have explored hundreds of technologies before settling on our current setup. We have talked with dozens of our users to understand and answer some of their hardest questions in seconds where they would previously have spent years. We continue to work on features that are requested but not yet possible for us to deliver. We are making rapid progress.

There are very few companies doing what we do at The Data City. There are very few companies doing machine learning as quickly and efficiently on commodity hardware as we are doing at The Data City. I would bet quite a lot that no-one is doing both things at once.

9. Describe a time you've had conviction in something, and then changed your mind.

I once believed strongly in the value of evidence-based policy, large-scale randomised trials in economics, and giving more power to technocratic structures to apply what we learn. I have changed my mind.

While less sceptical of our ability to measure what is valuable in society than some I see that beyond the purest of sciences, observations and innovations from one place or one context often apply poorly to others. So while the invention of the white LED has led to more efficient lighting everywhere with almost no modification, the actions of large centralised institutions, even when those at the centre are brilliant, has led to poorer outcomes and less technical progress than in smaller or decentralised models with less brilliant experts.

Just because we cannot scale evidence-based policies and technical solutions to the whole world is no reason to abandon the tools we have developed. We can instead focus our efforts on smaller scales, though always ones at which large economies of scale can be achieved. We can lean more on the power of markets, especially citizens choosing which products they buy and which politicians they elect, to deliver better outcomes and faster technical progress.

Providing excellent transport in large British cities is primarily a technical challenge. The failure of Britain’s national institutions to apply their country and their capital's excellence to any other large city suggests that more local and more political approaches are required to achieve the technical step-change in human progress that we want, that we know is possible, and that we could be exporting around the world within five years.

10. In exceptional cases, we are open to experimenting with co-led programmes. Would you be interested in running an ARIA programme with someone else?

Not really.