Modelling Journeys and Transport Networks

Transport Systems Catapult

The average British driver spends 32 hours each year stuck in traffic, at a cost to the economy of £30.8bn. These numbers fail to convey the human price of parents racing on the school run, meetings and connections missed, products delivered late and noxious gas clouds swirling above queues of idling vehicles.

Britain is the fourth most congested country in the developed world. To the team at the Transport Systems Catapult (TSC) it seemed absurd that an advanced nation should be hostage to such an analogue problem in this digital age.
Smart phones are now location beacons and movement sensors; cars are Wi-Fi hot spots; computers can sift vast troves of data to identify hidden patterns. If these disparate capabilities could be harnessed, we might create a cure for congestion.

“If you’re running the roads you need to use simulations – realistic computer models – to see what effect building works, accidents or new designs might have,” says TSC Project Manager, Paul Bate. “Highways England has built excellent models but the files are so huge it took days to run a single simulation. That limited their usefulness.”

“We looked at technology we could bring in from other sectors, particularly on-line gaming, where thousands of people can be playing simultaneously, manipulating elements in a detailed virtual world, all reacting in real time. That’s when we called Improbable.”

Restless curiosity and a shared love of computer games prompted university friends Herman Narula and Rob Whitehead to start software business, Improbable, in 2012. They created an online platform for gaming, Spatial Ordinance Survey, which could simulate thousands of simultaneous actions in a virtual universe. TSC knew the system’s capability having trialled it some time before on a small-scale project. When Paul Bate called this time, it was with an altogether more challenging brief.

TSC, Highways England and Improbable set about reproducing the motorway network in digital form, complete with millions of independently acting vehicles.

As the prototype system came together, the team added new sources of information that would make the simulation even more accurate: intelligence from the trunk roads feeding onto motorways; then, live data from a telecoms provider. For the first time the team could see where travellers were coming from before they hit the highways. This was the breakthrough.

“The Transport Systems Catapult was really proactive and hands on”, says Aleksandra Laska, Improbable’s New Business Director. “They got lots of smart people together, all with different skill sets, to brainstorm and share ideas. Every two weeks we would have what we called ‘sprint reviews’ to see if we were on the right track and the whole project moved quickly.”

“The speed of development was unprecedented”, Bate admits. “But the important thing is: it works as planned and we’ve done it at scale. We can now get near instant answers to ‘what if’ questions. It is a wake-up call for a conservative sector and could change the way people run their networks.”

With the demonstrator now doing the rounds at seminars and trades shows, Highways England and Improbable are working on plans to develop the technology over the next two years. TSC, meanwhile, is showcasing the opportunity to other parts of the transport industry.

Bate adds: “Nothing like this exists for the rail network which, in some areas, is still reliant technology from the 1960s. Much of their knowledge is localised with people only able to see what’s going on in their area. That means a signal operator in Norwich might send a freight train up the line, oblivious to the fact that it will arrive at Manchester
Piccadilly just in time for the evening rush hour.”

“In 2015 we commissioned The Traveller’s Needs Study, a huge piece of research involving 10,000 people. It identified that 75% of journeys involved ‘pain’ – by which we mean hassle and unnecessary stress, largely because of the different ways of travelling – car, bus, train, bike – don’t join up.

“Ultimately, we want to be able to model any journey, anywhere, combining every form of transport and show
it as one joined-up system. Companies can then mine this resource to create products that make it easier for
people to get about, like a single ticket that works across all types of transport or timetables that reflect the
different phases of a journey. If we can do that, it would be truly transformational, a world-first.”

Contact us