The need to control road traffic flow is nothing new. Indeed, the world’s first traffic light appeared in London ­– at the junction of George and Bridge Streets, near the House of Commons – in December 1868. Though this fledgling effort to improve the lot of the city’s road-bound public obviously pre-dated the motor car, as the commuting paradigm inexorably shifted away from horse-drawn carriages, the demand for better traffic management grew exponentially.

Over the years there have been many attempts to model vehicle flow mathematically, ranging from the – not entirely successful – application of methodologies derived from fluid dynamics, to the development of classical two-phase, and even three-phase, traffic theory. However, such theoretical models often correlate poorly with observed events.

Since commuter experience is inevitably linked to the hard realities of the road, the case for enhanced observational infrastructure and novel applications of monitoring technologies largely makes itself. Just as the transistor radio’s rise to ubiquity and acceptance in the 1960s ultimately paved the way for the in-car traffic updates we now regard as routine, the current crop of internet-connected portable devices, sat-navs and GPS-enabled systems have opened the door on the next stage. If the trend continues in the way many believe, the rise of digital monitoring promises a real-time revolution for road users everywhere.

Joined up thinking

"We’ve all got used to being able to plan our journeys via GPS, our phones and computers, and finding out about road works in advance, but that was just the ‘phase one’ stuff" explains recently retired local authority transport engineer, John Ritchie. "Now it’s all becoming a lot more joined up. We’ve been talking about intelligent transportation systems for years – the difference is they are only just beginning to be intelligent enough to justify the name. Like everything else, it’s the networkability that makes the difference."

“Effective integration is key to making dynamic traffic flow updates available to motorists.”

Effective integration is key to making dynamic updates available to motorists. This calls for a range of individual synchronous elements, from the traffic enumerators and convergence indexing systems, to the variable message signs along the carriageway. If such functional synergy has its benefits now, for developments like the much heralded road-train "platooning" on automated highways to become future reality, it will prove essential. Many aspects of the necessary technology are, however, already here, and other parts are tantalisingly close.

Active traffic management

The active traffic management (ATM) of the UK’s M42 motorway is a case in point. Motorway incident detection and automatic signalling (MIDAS) sensors permanently installed at 100m intervals monitor traffic flow, with specialist software analysing their data to determine appropriate speed limits, which are displayed on gantry-mounted signs. In addition to this computerised system, human operators monitor 150 CCTV cameras along the ATM-controlled section of motorway, and are able to communicate as necessary with drivers via a series of variable message signs along the route.

A Highways Agency survey found that nearly 70% felt better informed about traffic conditions as a result. The results also show near universal compliance with the variable mandatory speed limits. No doubt the automatic number plate recognition and digital enforcement cameras which also feature in the system undoubtedly played their part.

Although some aspects of the scheme – notably the opening of the hard shoulder to provide an additional lane – provoked some controversy, there is little doubt that as a means of promoting optimal traffic flow on congestion-prone routes, reactive real-time monitoring works. However, it still remains a ‘top-down’ management approach; things become even more interesting when connected drivers are empowered to use the information for themselves.


The Highways Agency, for instance, which is responsible for England’s trunk road and motorway network outside of London, is one of a growing number of organisations to make traffic information available online, via RSS, desktop applications, camera feeds, email alerts and Vista / Windows 7 widgets. Traffig Cymru and Traffic Scotland do much the same for their respective parts of the UK – and all three have recognised that for people on the move, being connected is not all about websites and PCs, so are dedicated to providing facilities for PDAs and WAP devices.

Communication hubs

Disseminating the information is, of course, only part of the process and there is much unseen ICT infrastructure between the roadside and the display screen. For example, around five miles of copper wire interconnects every group of eight data-gathering cabinets in the Washington State Department of Transportation (WSDOT) Pudget Sound area monitoring system. Each field cabinet is equipped with a Bell type 202 1,200bps modem, the communication hubs that serve them are located at ten-mile intervals, with extensive fibre-optic cable running between hubs and the traffic management centre. This, however, pales into insignificance when you compare the hardware required to take images from the systems’ extensive array of monitoring cameras available online.

“Human operators monitor 150 CCTV cameras along the ATM-controlled section of motorway.”

Sharing video signals

Sharing the same fibre-optic backbone, video signals from many individual cameras at the communication hubs are combined by frequency division multiplexing. This is an analogue process that allows 16 CCTV streams to be transmitted up to 35 miles via a single optical fibre. At WSDOT’s traffic management centre, these signals are de-multiplexed and connected to a 448-input / 112-output port video switch controlled by a Windows server. In turn, this server is controlled by operator consoles that allows output to display any camera or sequence of cameras.

It is a triumph of interconnectivity and compatibility, resulting in seamless integration and flexibility thanks to cameras made by COHU, pan / tilts by Pelco, an American Dynamics video switch and a transmission system manufactured by Catel.

Video and data servers

Making the transition from the department’s intranet to the internet requires still more ingenuity. One of the output ports is connected through a Matrox Meteor video capture card to a dual 800 MHz Intel Pentium III running Microsoft Windows NT 4.0 Workstation, which acts as a dedicated video and data server (VDS). VDS multithreaded, multitasking software that was purpose developed in-house, creates a virtual emulation of one of the operator control consoles, directing the video switch to display a specific camera, the programme then capturing a single frame from the live feed. This image is then first saved as a device independent bitmap (.dib) before being read back and compressed into internet-standard jpeg format. It is then written back to disc and queued for web server access.

When an outside user selects the appropriate camera, the page displayed retrieves the appropriate jpeg file. Since the VDS programme constantly updates the web server content, the image appears as if in real-time, auto-refreshing every 90 seconds using an HTML client-side pull.

Clearly the technology and data infrastructure will have to become even more complex if many of the dreams for future systems are to be realised, but the benefits are potentially enormous. According to a 2009 BBC estimate, drivers listening to traffic broadcasts have a 15% chance of hearing something relevant to their journey. If digital services really do become as big as predicted, it might well approach 100%, at least on the major commuter arteries. A development that could change our experience of road travel forever.