Home / Features / Intelligent uses of big data for intelligent transport

Intelligent uses of big data for intelligent transport

Big data offers low cost, rapid, and real-time intelligence for the UK’s ITS sector. INRIX Chief Economist, Professor Graham Cookson, presents four use cases for big data focused around the UK government’s recent road investments plan.

Over the past fifty years the volume of traffic in the UK has risen dramatically – from 70 billion vehicle miles travelled per year in 1960-61 to 304 billion in 2011-12. Combined with sustained economic growth and increased efforts by cities to enable multi-modal transport options, road traffic may continue to increase by up to 50%. However, the level of public investment in the road network has not kept pace with this trend, with public spending a third of what it was in the 1970s and 80s.

As a result, congestion has risen dramatically, which has taken its toll on the economy. In fact, congestion imposes a huge cost on all sectors of the economy. The recent INRIX UK Traffic Scorecard estimated congestion cost UK drivers almost £31 billion in 2016, or around £1000 each. Businesses are not immune either, with the Confederation of British Industry (CBI) claiming that congestion is a major concern for 73 per cent of UK businesses.

In response, the UK has announced the largest road building programme since the 1970s, tripling the annual investment in major road schemes by 2020-21. In 2013, the UK government committed £11.4 billion of capital funding to improving motorways and A-roads by 2020 under the Road Investment Strategy, and in November 2016 the Chancellor announced a further £1.3 billion to tackle congestion through the National Productivity Investment Fund (NPIF).

INRIX1-1707

Plans include adding extra lanes to the busiest motorways, identifying solutions to tackle some of the most notorious and longstanding traffic hotspots in the country, upgrading the national non-motorway network managed by Highways England to dual-lane, and £10 billion to repair the national and local road networks.

However, there is what economists call an opportunity cost involved in each decision. Every £1 invested in a specific transport project is £1 that isn’t spent improving transport elsewhere in the network. It is therefore critical that we maximise the impact of our road infrastructure investments and ensure that we are targeting the worst affected areas. Furthermore, to aid accountability and to inform future decision making it’s critical that we evaluate the performance of these improvement programmes.

This is where big data can help. Today vast quantities of data are generated and stored every day on the transport network. From 20 million daily ‘taps’ with an Oyster travelcard or contactless bank card on the London Underground network to millions of cars generating GPS data. Converting the data to insight is the challenge and requires both technical skills and sectoral understanding. To illustrate how to leverage big data to provide intelligent insights for the ITS sector, let’s consider four use cases.

Big Data for Rapid Project Appraisal
Concerns have been raised by the UK Government’s National Audit Office (NAO), about whether the Road Investment Strategy offers value for money or is even feasible by 2020. The NAO said: “The Department [of Transport] and Highways England need to agree a more realistic and affordable plan if they are to provide optimal value from the Road Investment Strategy”.

One of the NAO’s main concerns was the speed with which the UK Department for Transport performed the initial project appraisals at the expense of value for money. But the UK government at both central and local levels can now conduct project appraisals and post implementation evaluations in record time by leveraging the power of big data, and on-demand analytical and data services, such as INRIX Roadway Analytics. There’s no excuse for not acting quickly as there are cost effective, cloud-based products on the market that use big data to help solve problems like these.

Big Data for Optimising Road Investment
The Chancellor ring-fenced £1.3 billion of the National Productivity Investment Fund, announced in the Autumn Statement, for roads. The UK’s strategic road network will receive £220 million of this pot and Transport Secretary Chris Grayling recently announced which traffic hotspots will receive the first £110 million. But has the Department for Transport identified the best schemes? Is the taxpayer getting value for money?

As featured on Sky News and BBC News, November’s INRIX “UK Traffic Hotspots” study analysed every traffic jam across 21 UK cities and identified over 20,000 recurring traffic hotspots. Using this rich dataset, INRIX ranked the UK’s worst traffic hotspots by their total impact on car drivers and passengers.

INRIX2-1707

It’s great to see that the UK government’s latest announcement will tackle a number of these hotspots, including the A63 in North Yorkshire, the M5 around Bristol, the M25 near Heathrow airport and the M27 to the south coast. However, a number of the UK’s worst traffic hotspots are omitted from the government’s list, including the A48 in Cardiff, the A720 outside Edinburgh and the M62 in Liverpool, which collectively had 1920 jams last year according to INRIX Roadway Analytics.

Ranking traffic hotspots by total driver impact is quick and simple using on-demand, big data analytics, and combined with cost-benefit calculations it can be used to optimize road investment to maximize the total benefits to drivers and the wider economy.

Big Data for Intelligent Transport Systems
The use of real-time traffic data in intelligent transport systems is not new. However, big data generated and stored remotely allows transport planners and network operators to offer dynamic, adaptive managed transport solutions that can be deployed at relatively low cost on a wide scale.

Traditional infrastructure based solutions, for example traffic loops, are not cost effective when deployed at scale, and cannot adapt to the ever-changing needs of network operators. For example, it is not feasible for a strategic road operator to install and operate infrastructure in neighbouring feeder roads, yet intelligence of this sort may help operators manage strategic road networks more efficiently in real time.

Take traffic light optimisation as an example. Big data has two main advantages over more traditional approaches. First, signals can be optimised in real-time reflecting the actual road conditions by the minute. Second, the marginal cost of adding an additional ‘smart’ signal can be significantly higher when using hardware rather than a data based approach.

Our recent partnership with the City of York is a great example of how this works in practice. During the City of York project, INRIX real-time traffic data will be used to optimise traffic light phasing along the heavily congested A59 corridor.

Big Data for Post-hoc Evaluation
Post implementation evaluation is now standard practice across the transport sector. However, it can be costly and time consuming and requires adequate baseline data to enable an accurate evaluation.

A big data approach to post-hoc evaluation has three main advantages: speed, cost and convenience. First, as the data are collected and stored in real-time the analysis can occur as soon as evaluators want after the project starts, and the data can be accessed immediately. Second, the costs of accessing and analysing traffic data from big data sources is significantly cheaper than a bespoke data collection exercise that most agencies currently use. And finally, as data are collected and stored automatically across the road network, there is no need to consider and specify a data collection exercise to form a baseline. Data can be retrieved at any point in the project cycle.

As an example, consider this evaluation of Munich’s Mittlere Ring. In the INRIX 2015 Traffic Scorecard, it was Germany’s most congested road, wasting drivers as much as four days a year. In response, the Government of Upper Bavaria committed Euro 400 million on a range of infrastructure projects to tackle this problem., including the 1.5 km Luise-Kiesselbach-Platz tunnel that opened in July 2015. Using INRIX Roadway Analytics we were able to look at the impact of the tunnel on the ring road traffic by comparing average traffic speeds before (October 2014) and after (October 2015) it opened to the public. There was an approximately 10 kph increase in both the AM and PM peak hour speeds – the tunnel clearly worked.

INRIX3-1707

Conclusion
This kind of big data analytics should be at the centre of project design, project evaluation and decision making in 2017. It can also be leveraged in the active management and optimisation of the road network. However it is used, big data analytics offers a number of advantages over traditional methods and techniques including speed, cost, scalability, adaptability and offering real-time insights. ◆

p-INRIX-Cookson-1707About the author

Professor Graham Cookson is the Chief Economist (EMEA) of INRIX and a visiting professor of economics at Surrey Business School, University of Surrey.

He joined Inrix in 2016 and leads the company’s industry-leading economic research team in Europe, a recognised voice of impartial, data-driven analysis of transportation and connected car services.
Graham and the INRIX Research team leverages INRIX’s 500 Terabytes of data from 275 million different sources covering over 5 million miles of road to produce valuable and actionable insights for policy makers, transport professionals and drivers.

In addition, he leads on the development of INRIX’s industry-leading annual Global Traffic Scorecard. Graham has presented to various government agencies and at numerous industry conferences, and has been widely quoted in national and local media.

Prior to joining INRIX, Graham was Chair of Department and Professor of Economic & Public Policy at the University of Surrey (UK).

Graham has a Bachelor of Arts in Philosophy, Politics & Economics from the University of Oxford, a Postgraduate Certificate from King’s College London and a M.Sc. and Ph.D. in Econometrics from Imperial College London. He is fellow of the Royal Statistical Society and a member of the Royal Economic Society. Follow Graham on Twitter @grahamcookson and on LinkedIN at www.linkedin.com/in/gcookson