Swarm Intelligence to curb Traffic cognation
Over the last decades vehicle populations have significantly increased worldwide. Such heavy growth in the number of vehicles has caused massive traffic congestion resulting high probability of road accidents. According to a data of Department of Transport Management (DoTM) the total number of motor vehicles running across Nepal has reached nearly 3.1 million as of mid-May. The department registered 1,691 minibuses, 957 mini-trucks; 3,966 units of heavy equipment, 7,651 units of cars, jeeps and vans, 4,154 pickup vans, 10,076 three-wheelers, 6,189 power tillers and 6,131 e-rickshaws as of mid-May of Fiscal Year 2017/18. Similarly, 183,217 new motorcycles were registered during the period.
In such context, traditional control mechanisms are not able to provide sustainable solutions for Traffic cognation because the flow of vehicle in Kathmandu is always non-linear and it requires high intelligence system to understand the traffic pattern. Also the traffic condition changes on hourly basis as early morning traffic and Day time vehicle flow pattern is very different. So the control tactics should be highly responsive. Due to drastic uprise in traffic networks and haphazard vehicle culture; it seems complicated to make a real-time implementation of current traffic control model in Kathmandu. Therefore Nepal should experiment with new and very innovative traffic solution method, at lease inside valley it should be done in no time.
If we observe the colony of ants marching in and out of a nest, we might remember the highway bustling with traffic. From Purano buspark to New Baneswor which is maximum 15 minutes’ drive, people spend not less than seven minutes on average in congested traffic. From Newbaneswor to Balkumari also same doomed experience will reoccur. But you will never see ants stuck in such congestion despite thousands walk together in narrow trail. If we all drove more like ants walk, there would experience fewer traffic jams. Ants, termites and other swarms have something in common that human don’t- Unlike swarms human drivers never put aside their own interests to make smooth traffic happen. Some researcher has even suggested implementing insect’s mobility algorithm in human road behaviour to ensure sound traffic like swarms.
Swarm Intelligence (SI) is a subfield of Computational Intelligence related with the evolution of bio-inspired intelligent systems. SI approach uses the collective mobility behavior of natural agents such as colony of ants and flocks of birds as an inspiration to re-create their guiding algorithms. Those algorithms have been proven as efficient mechanism in solving real-world problems. Scientifically speaking, to solve real world traffic problems using swarm intelligence algorithms, we need a mathematical representation of our traffic cognation data, which would, describes our problem and all the decision variables leading to the problem. Then we can graph the optimal solution mechanism accordingly.
In fact, Ant colony optimization (ACO) algorithm has already been applied to solve traffic related problems, such as vehicle routing problem and traffic cognation. Ant-based systems have special properties such as adaptability, dynamicity and scalability which are the main requirements for solving vehicle traffic congestion. Cars also move like ants, they leave pheromones in terms of perceptional signals; Braking lights, Turning lights and Changes in speed. One moving car sniff pheromones released by other cars and adjust their speed and direction accordingly.
Dr. Couzin, a mathematical biologist at the University of Oxford studied army ants as well as other swarming animals and discover simple rules that allow swarms to work so well. Deciphering those rules was big challenge because the behavior of swarms emerges randomly from the actions of thousands individual ants. To understand the pattern of swarms he builds a computer model of virtual swarms. He found that each virtual ant leaves a chemical marker that attracted other ants while the marker remains still fresh. Each ant could sweep the air with its antennas; if the air made contact with another ant, it would turn away or slow down to avoid collision.
In the case of army ants returning to their nest with food in a dense column, incoming lane is bordered by two lanes of outgoing traffic thereby making three-lane highway to avoid crash. Are human different from army ants? Pay attention next time in busy crossroad, we naturally form similar lanes in crowd while walking. But whenever we get on wheels, we think more about minimizing our travel time and not consider other moving vehicle. We would change the lanes; sneak our two wheels wherever we see space. It’s also discovered that ants impulsively form highways. If the ants going in one direction become dense, their chemical trails attracted more ants heading same way. This response directs other ants to form a single packed column. The ants going the other direction will turn away from the oncoming traffic and form adjoining lanes. In short, the secret to such robust mobility management in ants was because of cooperation and strong sense of common interest.
Similarly a Physicist Apoorva Nagar also believes that there are three main reasons ants don’t jam up. Firstly ants don’t have egos and they don’t show off by zooming past people. Secondly they do not mind a few accidents or collisions unless there’s a serious pileup, they just keep moving. Thirdly, ants seem to get more disciplined when paths get crowded — something which humans are notoriously bad at. They’re less likely to make unexpected moves in such sort of heavy traffic.
To reduce vehicle delay at intersection, easy routing and avoid the traffic congestion in Kathmandu, ant model traffic behaviour seems an effective way. There has been many detailed research about Swarm intelligence in traffic behaviour regarding; how to get precise traffic frequency, the way to module traffic signals and maintain systematic traffic behaviour. If Nepal government consider such innovative traffic research and implement it in Kathmandu roads, then it surely will minimize the degree of traffic cognation we are experiencing today.
Refrences:
Apoorva Nagar:
https://io9.gizmodo.com/what-ants-can-teach-us-about-avoiding-traffic-jams-1681565109
Comments
Post a Comment