Literally overnight, an information technology (I.T.) system completely changed how 100,000 people use their cars. No, this is not some new telematics system. The I.T. wonder is London, England’s new Congestion Zone Charging. Launched in February, 2003, the system detects all vehicles that are driving in the city’s center, then charges them five British pounds (about $8 U.S.) for doing so. There are no toll booths. Instead, there are 688 pole-mounted cameras that scan vehicle-license plates. Image-recognition software identifies the plate number on each vehicle. By 10 pm that day, the vehicle driver must pay the five-pound fee or get fined 40 pounds for non-payment. Several American cities including the municipalities of New York, Boston, Philadelphia and Chicago are seriously looking at the British system.
The project’s technology, politics and launch are fascinating. London’s socialist mayor, Ken Livingstone, immediately championed the idea upon being elected three years ago. A lightening-fast system implementation was essential to minimize the enormous pressure to scuttle the controversial project. Derek Turner, Managing Director of Transport for London (TfL) Street Management, employed a fast-track procurement process to line up all the vendors and key technology rapidly.
Capita Group, an EDS-like company based in the UK, was the lead systems integrator. Capita is also the current operator of the system, managing the payment system and dunning operations. About 500 staff the system. Deloitte Consulting also played a critical role. In addition, London’s TfL was instrumental in specifying two key components: the cameras and the license plate-recognition software. What’s impressive about the $200-million system is its ability to identify vehicles automatically as they ordinarily move about the city.
With regard to the technology deployed, there is a mix of black-and-white and color cameras. Both types use “X-wave technology.” This enables them to see better in poor light conditions (in the rain, at night, etc.). They feed compressed images via fiber-optic cable back to the system hub, where the actual recognition takes place.
In addition to the image-recognition equipment is the I.T. payment software. Drivers pay the five-pound fee principally through one of 200 retail stores. Also popular is paying over the phone using a credit card. The I.T. payment system handles about 100,000 transactions/day.
Each night an Oracle, database-management system matches that day’s payment information against the license-plate numbers collected. Those who did not pay must then be identified and fined. This is done by extracting vehicle-registration data from England’s central Driver Vehicle Licensing Agency (DVLA). Pulled from the DVLA are both the owner information as well as the make and color of the “guilty” vehicle.
Capita Group employees manually inspect the color, street-scene photo of the “guilty” vehicle. They do so to make sure the photographed car model and color correspond to the DVLA-stored model and color information for that license plate. This is necessary since the license-plate recognition software is only 90% accurate. About 10% of the vehicles entering the Congestion Zone get fined each day.
Other cities have employed congestion-charging schemes, but on a much smaller scale and with more intrusive technology than London’s. For instance, Singapore requires that every vehicle entering the city have an electronic “tag” installed. The vehicle then must drive in a restricted toll-booth lane that “reads” the tag. The sheer number of vehicles and geographical size of London far surpasses that of other cities doing congestion charging. TfL’s Turner acknowledges that the fiber-optic network is a limiting factor in London’s own I.T. system.
Not only was the implementation of the system remarkably fast—going from idea to operation in less than 2.5 years—but it seems to be working, lessening congestion. The number of vehicles in London’s city center is down 17%. The traffic is even moving faster than the average 9 mph that it was crawling at prior to implementation.