Business impact Reduced traffic congestion in Stockholm by 25 percent during peak business hours, while concurrently increasing mass transit usage by 40,000 riders per day.
Issue Stockholm wanted to reduce both the number of traffic jams in the city and the municipality’s air pollution levels. To accomplish this, Stockholm’s government decided to reduce traffic by directly charging drivers who use center city roads during peak business hours. The City of Stockholm and the Swedish Road Administration engaged IBM Global Business Services to design, implement and operate a trial road-use charging system. The solution includes a network of 18 drive-through control points surrounding a 24-square-kilometer area of the city center, plus the central computing systems for capturing vehicle identification data and billing drivers for road use.
Many drivers, at the city’s urging, retrofit their vehicles with Radio Frequency Identification (RFID) transponders, which communicate with receivers at the control points and trigger automatic payment of road use fees. Cars passing through the control point are also photographed, and the license plate numbers are used to identify those vehicles without tags and to provide evidence to support the enforcement of non-payers.
Many of the photographs are quite clear and the license plates can easily be recognized by an automated optical character recognition (OCR) system. The system records the number and then goes on to process the payment or billing tasks. When nighttime conditions, dirt or bad weather come into play, automatic recognition becomes more difficult. Operators are required to check the unclear photographs and manually enter the license numbers into the system. During preliminary trials the system was only able to automatically identify a small portion of the photographed cars using existing optical recognition technologies.
Executive summary Road-use charging systems must demonstrate a high degree of accuracy. An inefficient system costs the city funds it would otherwise have earned from tolls. Incorrect billings also cause bad feelings among residents. Understanding this, the IBM Global Business Services team called upon IBM Research to help boost the accuracy of the automated license plate recognition system. IBM researchers delivered an OCR engine that significantly increased the number of license plates automatically identified in good weather — with still greater improvement anticipated during bad weather. This translates directly to a significant reduction in manual data entry costs.
What IBM did Because of varying degrees of illumination, bad weather and sometimes-awkward camera angles, not all license plates photographed by cameras at the roadside control points can be automatically identified by standard OCR systems. IBM Research, calling upon deep industry expertise and a fair bit of ingenuity, developed a sophisticated "level 2" recognition system that employs sophisticated algorithms to make a second attempt at identifying unclear license plate images.
These algorithms use techniques such as image enhancement and comparison of front and back license plates to analyze the entire image and search for predefined patterns. Mimicking the human eye, the algorithms decipher images of often barely legible text by moving the image around until an optimal viewing angle is found, and the expected pattern can be recognized. After recognition, the IBM system automatically records the license plate number, matches it with a vehicle registry database and processes billing. Drivers can pay their bills at local banks, over the Internet and at area convenience stores.
Capabilities applied Drawing upon its experience in Singapore (where IBM took part in the development of the world’s first electronic road pricing system), along with relevant work in Canada, the United Kingdom and Austria, IBM was well positioned to share with Stockholm its expertise in developing and executing a road-use charging system. The Stockholm solution represents a new benchmark in the scale, scope and sophistication of the technology underlying the city's innovative approach to addressing traffic congestion and related environmental problems.
| |
