Welcome Your IET account
Steve Martin and John Candy, in a promotional photo from Planes Trains and Automobiles

AI and the great Christmas getaway!

Image credit: Rex Features

Artificial intelligence may be the key to improving traffic management during the busiest time of the year.

At a time when we’re all trying to get away for a Christmas break, the very last thing we need is a train out of service or a bottleneck on a motorway. Changes in expectations - what is deemed acceptable and not acceptable - mean that reliability and a smooth ride is of paramount importance for service providers and passengers alike. This is even more so in areas where the passenger pays for the service.

Expectation has dovetailed with smarter technology. This allows for better management of traffic flows and more effective and efficient maintenance programmes to significantly improve the availability of the infrastructure upon which we rely.

This is good news. Infrastructure generally is, we know, under pressure. According to the World Health Organisation (WHO), 50 per cent of the world’s population now live in cities and this figure is growing by about 2 per cent each year. This means that cities need reliable and fast means to get around both within the city and to and from it.

How, then, has the use of technology improved traffic flows and reliability of infrastructure?

When it comes to planes, Rolls-Royce - a leading manufacturer of aircraft engines - is using Azure Stream Analytics and Power BI to monitor data that comes from sensors attached to each aircraft along with more contextual information such as air traffic control, route data, weather and fuel usage. The idea is that as well as having a standard maintenance programme the company can use the data to get a better real-time picture of the health of its aircraft engines and use it to help with unscheduled maintenance, aircraft choice and route selection.

Nick Ward, head of OEM digital solutions at Rolls-Royce Digital, explains: “We have what we call an ‘intelligent engine’. While operational, each aircraft sends us data and if we need to do something to a component while on a layover, for example, we know in advance and can have the people and the parts ready for when the aircraft lands. It’s picking up anything that is non-standard so that the schedule is not disrupted.”

He says that today’s passengers are less likely to tolerate delays and so the onus is on the airline and the aircraft manufacturers to make sure that the aircraft is ready to go on time.

“The goal really is to make sure that every Rolls-Royce-powered aircraft takes off and completes its mission as planned in the most efficient way possible. So, it is foremost about ensuring the availability of the aircraft. But we also use the data to inform flight times, fuel efficiency and general operational performance,” says Ward.

For trains, the concept of reliability and efficiency has been harnessed by Siemens within its train rolling stock. Like Rolls-Royce, it uses sensors and data analytics to turn its maintenance programme from reactive to proactive. Sensors feed back information on trains and tracks; the data is then run through diagnostic analytics to reveal if and when a failure is likely to happen. By monitoring information in near-real-time, it can quickly react to concerns before they disrupt services. If an anomaly is detected, the component is sent for inspection. This helped the company move railway maintenance methods from reactive to proactive

Jo Binstead, head of innovation at Siemens, says: “We’re really working on remote condition-monitoring and predicting maintenance issues to make sure our rolling stock is available and reliable. We’re also a safety-critical industry so the more pinpoint accuracy we can get the better.”

However, she points out that the accuracy of proactive maintenance is highly dependent on how old the stock is. An aged fleet cannot provide as many data points as newer stock but, consequently, as stock is modernised the accuracy and efficiency of proactive maintenance will increase. “New fleet, like the Thameslink trains, now have data on sub-components that allow us to know in real-time what is actually happening and to take steps to remedy any issues,” she says.

The transportation issue is not just about maintenance. It is also about capacity and ensuring that transportation can flow as effectively as possible given the increased strain on it. On the railway network, one way of doing this is by automating signalling and monitoring how long a train is sitting in a station and how to improve flows. Measures like this can increase capacity by up to one-third, says Binstead.

Nowhere is capacity more of an issue than with the road network. Indeed, the UK is ranked the worst in Western Europe for traffic. Drivers in the UK lose about £1,200 in lost productivity, wasted fuel and time in the average of 31 hours spent stuck in traffic jams, according to data from traffic data company Inrix. The cost to the economy, it says, is somewhere in the region of £37.7bn a year.

One attempt to deal with this is via the smart motorways system. It uses radars and loop detectors to measure both speed and vehicle flow and uses this data to set appropriate speed limits, thus ensuring that the traffic actually flows rather than becomes stop/start. In some cases, the hard shoulder can be opened to facilitate flows.

Jo White, head of intelligent transport systems at Highways England, comments: “We have a range of options from setting a lower speed limit to opening up the hard shoulder to all-lane running, which is when we convert the hard shoulder to a lane. We also need the data to tell us where the back of the slowed traffic is and set speed limits approaching that to protect those that are slowed and to encourage an overall flow rather than sharp braking.”

She explains that the system takes real-time data and applies it to underlying algorithms to detect what the actual level of congestion is and set the most appropriate speed limit. It also uses historical flow information to detect whether there is something abnormal happening, like an accident, as opposed to just heavy traffic, and then escalate that information to the National Traffic Information service, which can then inform people as to what the likely delay is.

London has similar technology, called Scoot, which uses data from sensors within traffic lights to monitor flow and calibrate the lights to keep the traffic moving.

IBM, meanwhile, has been granted a patent for an AI-powered traffic-management system. Here the idea is to use cameras to provide real-time footage of traffic in each direction and then calculate the optimum traffic pattern. This is even more valuable in areas where it is hard to define a typical day. For example, a large group of children may be waiting to cross the road, or a backlog of vehicles is building up because they can’t turn across the path of oncoming vehicles, or a car is stopped on a red light in a deserted street; with this system these could be more effectively and quickly dealt with.

Technology, then, can help, and the aim is not just to improve flows and availability on specific modes of transport – the new landscape is about a more holistic approach with end-to-end connectivity now being looked at.

Grant Klein, a partner at PWC, says: “Individual interventions are great, but it now needs to all be linked up to provide a seamless door-to-door experience.”

Within the train network, for example, this might mean better management of passenger flows within the station by, say, redirecting people to another escalator. This is something that Siemens is actively working on with the University of Sheffield. Its Siemens Sheffield Advanced Multimodal Simulator (S2AM) project uses micro-simulation techniques to provide a framework to model individuals as they move through an integrated, multi-modal journey.

The framework allows different simulations to be ‘joined up’, enabling complex modelling of the way that people make their journeys. It also allows for passenger preferences and for reaction to different situations such as crowded trains or other disruptions.

Another report along the same lines by Highways England looks at ‘mobility as a service’. The report, ‘Connecting the country, planning for the long term’, says that mobility hubs, where people change their mode of transport such as park and rides, airports and train stations, are likely to see increased attention as the focus turns to door-to-door transportation.

White comments: “We think mobility as a service is really important: technology can make sure that the integration point can remain as optimised and free-flowing as possible. This all adds up to a better customer experience. We can connect the original mode to the hub and then to other parts of the journey too. For example trains every 90 seconds on the Victoria line means people no longer rush and crowd escalators in the station.”

Klein says that the idea is to use technology to devise an optimum point-to-point route and method of transport. “Data can now be analysed using algorithms to provide us with information on the best way to make an entire journey and the integral parts within that. This information must, though, be backed up with a solid customer experience, because ultimately they are no longer passengers – they are customers paying for a service,” he warns.

Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.

Recent articles

Info Message

We use cookies to give you the best online experience. Please let us know if you agree to all of these cookies.

Learn more about IET cookies and how to control them