Future Energy Sources 2.5.1.2 Improved Sequencing of Traffic Control

Following on from last weeks thread on reducing the severity of traffic congestion by varying the time people commenced and finished work, this week, I would like to look at another traffic related problem.

The very first set of traffic lights were installed outside the houses of parliament in England, in December 1868 and were based on railway signals in use at the time. It consisted of the usual semaphore signalins plus a gas powered lantern with coloured filters that could be positioned in a way that indicated which vehicles should stop and which should proceed. It was not automated requiring the presence of a policeman to operate and unfortunately it blew up after a couple of months of service, injuring the officer that was operating it.

There is considerable debate on where and when the first electric traffic lights were installed but the following is a summary of the major steps in their evolution.

1. The very first electric lights were purely manual systems and required the presence of a human to operator and in reality were no different to having a policeman directing traffic manually. They did however make it safer and more comfortable for the person directing the traffic.
2. The next step in the evolution was to replace the operator with a system that ran on a simple timed cycle. Whilst these did not require an operator they could not respond dynamically like those controlled by a human operator.
3. With the advent of detectors it was possible to put pauses in the timed sequences. These pauses would prevent the lights from cycling to a position that allowed traffic to flow from a direction where no traffic was present. This was a big improvement getting rid of the frustrating problem where real traffic was halted for non existent traffic. However the systems were not perfect an often did not detect small vehicles like bicycles, motorbikes etcetera.
4. Next up was trying to synchronize traffic lights so that they all went green in relation to the lights immediately around them. This worked well in places like one way city grids but lacked any sort of dynamic control.
5. The final major step was to coordinate the sequencing of traffic lights from a central control system. This, in conjunction with sensors in the road, historic traffic volumes, remote cameras and digital computers enabled dynamic coordination of the lights to maximize the flow of traffic. A later development of this system was to place much of the computing power in the actual traffic light control units and getting them to communicate with each other over a network.

Ideally a system would consist of nothing but coordinated traffic lights that were controlled from a sophisticated computer network capable of dynamically coordinating the lights in a way that minimized the delay to all vehicles, but, this is not a simple task and there are several problems that make this difficult to achieve.

1. 1. First off you have bi directional carriage ways, you can only coordinate the lights maximize the flow of traffic in one direction. Traffic flowing in the other direction will end up being stopped at intersection after intersection and be subjected to inordinate delays.
2. Even if you do coordinate the lights to allow blocks or pods of traffic to get a series of green lights, there are several things that can mess up the flow.
1. If any of the drivers in a block drives faster than the speed the lights are coordinated for, the next set of lights will be red when they arrive forcing them to stop. The delay that they now introduce in getting back up to speed then throws out the coordination and messes up the flow for the entire block of traffic.
2. Drivers that dawdle and drive slower that the speed the lights are coordinated for will have a similar effect messing up the timing for the subsequent block of traffic.
3. You can only coordinate lights in this way for traffic flowing in one direction. Traffic flowing in the opposite direction will experience considerable delays.
4. Breakdowns and equipment failure can completely destroy the sequencing and throw the flow of traffic in all direction into disarray.
3. Not all traffic lights are capable of gathering the data and being controlled from a central point and therefore can't be coordinated.
4. A failure of the central control system can and has brought the flow of traffic to a complete standstill.
5. Coordinated systems are expensive and become more expensive as the number of intersections they control increases.

However, with increasing volumes of traffic in all cities around the world and if this increase it to be accommodated far more sophisticated ways or coordinating traffic flow need to be found.

What can be done?

• Is it possible to improve the coordination of traffic lights to improve the flow of traffic or is it a lost cause doomed to failure because of non compliant drivers?
• What about decentralized control systems, where individual traffic light systems are controlled dynamically by a local computer that coordinates the flow of traffic by communicating with other traffic light controllers over a network?
• What about the increased delays that a coordinated system can cause to traffic flowing against the main flow of traffic?
• Is it possible to develop control algorithms that can be used to improve the flow of traffic or is what is currently available as good as it gets?
• What about the systems that give busses and the like priority. They allow the bus to go first only for it to stop just past the intersection to pick up and put down passengers. You end up with all the vehicles being condemned to the speed and stop start progress of the bus.
• Can a coordinated system be secure enough to prevent hackers or terrorists from causing gridlock, bringing a city along with all its emergency services to its knees?
• Can we afford no to develop highly sophisticated traffic control system?