Information Age: News, analysis & insight for IT & business leaders

The Highways Agency is increasingly using business analytics to orchestrate traffic flow in real time

The Highways Agency has a data analysis job like no other in the UK.

As an executive agency of the Department for Transport, it operates, maintains and upgrades the road network in England – 4,500 miles of motorways and trunk roads – with a mission to provide safe and reliable roads, and to keep motorists moving.

Some of the challenges are long term, others need immediate action. For example, faced with an accident on the M25 that closes all the lanes in one direction, it knows that the tailback will lengthen fast – at a mile a minute – and knows decisions on how to divert traffic or warn approaching drivers need to be taken swiftly.

Until recently, the ability to analyse such situations was distinctly historical in nature: vast amounts of data were collected – from both internal and external sources – but that information on a particular stretch of road, on accidents or patterns of traffic flow was assessed at in retrospect.

Now, however, the Agency is enhancing its capabilities with live information feeds and analysis that are already enabling it to react to incidents and optimise the flow of traffic in real time.

To do so it is extending a 2005 initiative, the Roads Information Framework (RIF) programme, which involves the use of business intelligence (BI) tools and the creation of a data warehouse, and which has already dramatically improved decision-making and reporting capabilities that were previously based more on expert judgement than deep data analysis.

The technology behind that core infrastructure is MicroStrategy’s business intelligence platform – a technology set currently accessed by between 50 to 100 users focused on monitoring journey times, reliability and safety on the road.

The Agency already has lots of raw information streaming in, in real time.

Buried in trunk roads and motorways around England are over a thousand ‘inductive loops’ which measure the number of vehicles passing over them. These monitors can also detect the speed of the vehicle and its type – whether it is a car, an HGV or a caravan.

In addition, the Agency has deployed about the same number of more intelligent data capture drones – so-called Midas loops – which not only gather critical traffic movement data, but can also react to it. If the traffic conditions cross a certain threshold – such as a wave of vehicles slowing to below 30mph – Midas loops can automatically link up to roadside message boards to warn drivers about queues and other potential problems.

Supplementing those, the Agency also buys in data from third-party suppliers who use cameras around the network to monitor traffic speed and flow.

All of that is fed into the RIFs data warehouse, along with the operational logs from control centres. That enables the analysts to match traffic flows to a particular accident, for example, so they can see what the impact was on traffic flow.

But its aim is to augment its use of business intelligence for that kind of historical analysis by drawing on real-time data for forecasting traffic conditions and managing the network.

“Historically, the data was being collected, but what we weren’t doing was actually making full use of it,”

“What we are looking to do now is to move the analysis forward to look at what-if trends,” says Denise Plumpton, director of information for the Agency, “what might happen and how we can best predict – and then prevent – some of the serious congestion incidents that occur.”

The analysis is throwing up some fundamental new understanding. For example, it has found that the cause of two-thirds of all congestion has nothing to do with accidents or roadworks, but simply results from the volume of vehicles trying to use the network. “That [insight] has been really helpful in enabling us to focus on where we are going to put our future investment,” says Plumpton. “Of course, there is a focus on managing roadworks better, but trying to encourage people to travel at different times and go by different routes can have a far bigger impact.”

The BI tools have also been instrumental in the analysis of the changing patters of breakdowns. “Even though cars are getting more mechanically reliable, we are still getting large numbers of breakdowns. What’s changed?,” asks Plumpton. “It appears that one of the unintended causes of the SatNav revolution is that people run out of fuel far more frequently than they used to. Perhaps their reliance on the SatNav means they don’t actually relate the instructions telling them to drive so many miles in one direction to the amount of fuel they might need to carry out those instructions. But certainly, the rise in [empty tank] breakdowns has risen exponentially over the past couple of years.”

Armed with that specific information, the Agency has launched a roadside and media campaign encouraging drivers to check their fuel more actively.

And this is not a trivial problem. The sight of a broken-down vehicle causes many drivers to slow down in order to see what is going on – and that in itself can cause congestion and accidents. “So the quicker we can move the breakdowns on the better for keeping traffic moving,” says Plumpton.

The agency did one piece of analysis on how long vehicle recovery was taking on a particular stretch of the M62 motorway near Leeds and Bradford. What it found was that it was taking on average almost an hour to remove these vehicles.

Assessing the impact of that, says Plumpton, the Agency arranged for free recovery to a safe place for vehicles breaking down along that stretch of road, cutting the average recovery time to 20 minutes.

“While you might observe that we are using taxpayers’ money to provide a free recovery service, the net impact is very much in favour of the UK economy because we are getting traffic moving.”

That kind of “off-the-wall” analysis has helped the Agency assess where it should best invest its funds. “That does not necessarily mean building another lane or putting more motorways in place. Rather we can look at the other steps that we could take with our existing networks to improve people’s journeys.” Plumpton adds.

The next stage in its role out of BI is to push the information out to more people in its control centres. “Our Intelligence Units are staffed with experts in analysing information, but there can be a bottleneck in getting that information out – particularly as we want to move to real-time analysis of what is actually happening with any incident, in our centres” says Plumpton.

Today, the main role of these seven control centres across England is to manage incidents: deploy traffic officers, contact the police. The information they have at the moment does not give them much of a view of the impact of their decisions.

The agency wants to give control staff the ability to drill down into the data so they can see, for example, the impact of deploying traffic officers to a certain incident. If that is not proving effective, they can put a diversion in place. “They will be able to make the traffic flow in real time, rather than looking at it two days later asking ‘Why did we have that 10 mile queue?’.”

Anyone stuck in a tailback that is growing at a mile a minute will appreciate that – albeit with little knowledge of why they are on the move again sooner.

Further reading

Corporate game theory
Advanced military simulation techniques could be the key to better business outcomes

Business Intelligence 08
Creating the intelligent enterprise. A report from Information Age's Business Intelligence 08 conference.

Information integrity
Strategies for dealing with shadow data systems. An Information Age Business Briefing

Find more stories Business Intelligence Briefing Room