How the Internet Of Things can help build Smart Traffic Control Systems

With the presence of a traffic detection mechanism, a traffic control system can be fine tuned to generate optimal green and red times for the current cycle. Several customizations can be incorporated into the timing logic of the previous section, to produce more precise results. Some of these customizations are mentioned below-

Adaptive Traffic Levels – Instead of maintaining hard coded values of high, medium, and low traffic levels in terms of the number of vehicles, the microcontroller can process the data of the past few weeks to decide the values of these levels based on the time of the day. For example, the microcontroller can calculate the minima, maxima, and mean of the traffic between 8 a.m. to 10 a.m. for the past 3 weeks to determine the values for low, high, and medium traffic for this time slot. These values might be significantly lower for the next time slot (11 a.m. to 1 p.m.) when the office rush hours are over.

This setting of traffic levels based on previously recorded data avoids a scenario where the traffic always falls in the High category  (e.g. 15 – 30 vehicles) as per the fixed levels, thereby producing the same green time for both 15 vehicles and 30 vehicles and making the adaptive timing redundant.

Adaptive traffic levels will ensure that when the traffic is always greater than 15, then 15 vehicles will correspond to the low level of traffic and 30 vehicles to a high level of traffic, and therefore 30 vehicles will get a higher green time than 15 vehicles. This case is explained in the diagram below.

This improvement does come at a cost though because the storage of past data will drastically increase the memory requirements of the system.

Adaptive Green Times – Further adaptation can be introduced by the run-time modification of the green time for a particular level by the microcontroller, instead of a predefined green time duration for each traffic level. This can be done by keeping track of the number of vehicles left in the lane after the green timer expires or by measuring the number of seconds in which the lane becomes empty for a particular green time.

For example, one could implement an algorithm where the green time for a particular level of traffic is incremented by 5 seconds every time more than 5 vehicles are left in the lane when the red light comes on, and is decremented by n seconds whenever the lane stays empty for the last n seconds of the green time. This will further reduce unnecessary wait time for other lanes while ensuring that most vehicles in a particular lane are able to leave the intersection after one wait cycle.

Ambulance Detection – Yet another feature that can be added to the system is the detection of ambulances and fire vans in a particular lane. These can be detected based on the standard text and symbols displayed on these vehicles, by adding a symbol/text detection mechanism to the image decoding program. This will allow the microcontroller to increase the green time of a particular lane such that the ambulance or fire van does not have to wait for another cycle before getting the clearance to move.

These are just some of the ways in which the traffic data provided by the sensor can be utilized to create more intelligent traffic systems. Of course, the timing algorithms can be optimised further by using better machine learning techniques and fuzzy logic based classification but these sophisticated algorithms will require even higher processing power and memory.

In the next section, we will discuss how the Internet of Things can help in further improving these intelligent traffic controllers by creating a smart grid comprising of neighbouring traffic intersections.