The primary order of business Wednesday was to take my search for the best signal control system global. One could be forgiven for thinking that since there are large cities all over the world, there must be other signal control systems that rival the best ones in the United States.
It turns out that there is some merit to this way of thinking, but it must be heavily nuanced. America is unique in that it has relatively young cities that were developed with cars in mind. Their hallmarks are heavy use of grids and obsessive traffic control systems; quite a contrast to the ancient citadels of Europe with their medieval layouts, ring roads and excessive use of roundabouts.
It stands to reason that traffic engineers in New York and Los Angeles don’t build adaptive signal systems with giant roundabouts in mind. Nonetheless it seems worth examining how other large foreign cities tackle this problem.
I started with our English speaking neighbors across the pond. According to the Directorate of Traffic Operations for London Streets, they have around 6,000 signals in London, 3,000 of which are purely locally controlled (not controlled by their TMC). Many of these are pedestrian controlled mid-block crossings. Nineteen percent of them are centrally controlled fixed time lights and 30 percent (2,000) are adaptive.
The adaptive ones are managed by a system called SCOOT (the largest SCOOT deployment in the world, I believe) and traffic is modeled with VISSIM. In some ways it seems like they are paying a price for being early adopters of this kind of technology in that things like their detection technology (can’t measure actual flow, and detection is all at the beginning of blocks–not near the signals) and communication infrastructure (simple switched copper) can’t be readily upgraded to new systems. London deserves a lot of credit for their innovations. They have heavily modified SCOOT to overcome its weaknesses, including its generally poor performance under heavy congestion and greatly shortening its update times. It has been labor intensive but they have also been reaping the benefits for over a decade.
It’s worth pointing out that this is why we should all push for open standards. London didn’t have the choice back then. They were on the bleeding edge of technology. Now that this technology is maturing, we have no reason to let contractors bully us into buying closed systems that can’t be easily upgraded. Of course, I took the opportunity to mention Sensys’ impressive new detectors, New York’s controllers and LA’s signal control software.
An interesting tidbit: London has roughly half the signals of New York and three times the CCTV cameras. Aside from the “if you’ve got nothing to hide then you shouldn’t mind us watching you” philosophy, it’s worth mentioning that since London really has no reliable way of measuring traffic flow with their detectors, they devote a lot of man power to watching all those cameras and tweaking SCOOT. SCOOT is also primarily deployed within their congestion pricing zone. The zone helps prevent traffic loads from exceeding SCOOT’s limits. I’m going to try and find out if they will release their modifications to SCOOT to the public.
As I continued by virtual tour of Europe, I came across TNO a Netherlands Research Organization. TNO runs under an interesting model, 20 percent government-funded and 80 percent by its business-to-business deals.
TNO has been developing its own adaptive technology called Network Manager though they themselves don’t bring such technology to market. I’m sending their researcher’s links to some American system in an effort to learn the differences between them and with any luck encourage their mutual improvement.
Tomorrow, I’m going to track down information on UTOPIA/SPOT a system deployed mostly in Italy, UTMS — Japan’s primary signal control system, and Australia’s SCATS system which I believe is similar to SCOOT but with detection at the intersection instead of in advance.
So much fun, so little time!