Traffic congestion in London poses significant issues. On top of the stress and inconvenience for road users, congestion in the city can cripple individual and business productivity. There’s the impact on the environment of all those idling vehicles. And as one of the most visited capital cities on the planet, roads are also a vital lifeline for London and the wider country.
Part of the reason there is so much traffic build-up is that London experiences 20,000 unplanned transport incidents yearly. Addressing those traffic incidents promptly has been a challenge for Transport for London (TfL), the authority responsible for London’s transport network of road, rail, and underground.
Exacerbating this issue is the fact that when an accident happens, each passing minute left unaddressed means traffic builds and gets exponentially worse. Swiftly addressing a traffic jam within seconds rather than minutes could save the city countless hours, speed up travel, and markedly reduce pollution, one of the city’s key strategic targets.
What if you could bring together real-time data on those roads and spot an incident before someone picks it up on CCTV? That’s precisely what Andy Emmonds, chief transport analyst at TfL, wanted to achieve—a way to harness real-time data on all those roads and ideally head off congestion problems before they begin.
Uncovering hidden relationships across billions of data connections
Ultimately, the solution for TfL lies in the ability to access and derive help from data analysis at scale and from a multitude of inputs. Many London journeys are private and multi-modal (involving a combination of driving, cycling, taking the train, and walking), for example, making them difficult to track.
TfL’s historical approach was to collect distinct data sets, which meant they could only answer a fraction of the questions the team wanted to ask. Despite amassing terabytes of data weekly, the siloed approach to storage and analysis prevented TfL from drawing meaningful insights based on the relationships between these diverse data sources.
Emmonds discovered that graph database technology, a type of business database designed to store and efficiently navigate highly interconnected data, provided an effective solution to overcome these limitations. TfL needed a way to uncover hidden relationships and patterns across billions of data connections to make the decisions needed to predict and handle traffic incidents. A knowledge graph, a type of graph database, excels at harmonising complex data and flexibly modelling real-world structures and their business logic, and so was the ideal technology to underpin a new traffic management support tool for TfL.
The TfL knowledge graph was used in conjunction with what’s known as ‘digital twin’ technology to create a virtual replication of the physical transportation network. This approach enabled the TfL team to simulate and test various traffic incident scenarios and their impact on congestion before implementing strategies in the real world, providing confidence in their effectiveness.
If you liked this content…
With the implementation of the knowledge graph, TfL can now seamlessly integrate and analyse all those previously disconnected yet valuable data sets within the digital twin environment.
To evaluate the effectiveness of this new solution and assess the real-time insights it could provide, TfL conducted a highly successful proof of concept (PoC). The PoC yielded such compelling results in a remarkably short timeframe that TfL promptly approved the full-scale deployment of the system.
New networks and new ways of thinking
TfL hopes its graph-powered digital twin will play a crucial role in its vision of cutting congestion by 10% a year.
Given that the Lord Mayor’s office believes congestion is costing the London economy £5.1 billion a year, that reduction in congestion would return £500m to the city’s economy and deliver the equivalent in saved productivity of over £1,200 per driver per year.
Emmonds also hopes to build an optimisation tool for peak traffic days e.g. when a big sporting or entertainment event takes place, to efficiently plan routes across the network leveraging data from the new digital twin.
Let’s leave the last word to Emmonds. “Ultimately, what we’ve created is a holistic product [that’s] enabled us to establish new networks, new ways of thinking, and unlock efficiencies.”
Those new ways of thinking are now cohering around aspirations to use the technology as a basis for a much more ambitious kind of model: not just traffic (important as that is), but a potential smart city model. “[In a smart city],” he says, “people can move around; at one minute you want to do some shopping, then you want to go on the transport network, and then you want to do something different, maybe you want to be entertained. We could move to making all that seamless.
“And that, I think, would be a great vision for London.”
