Optimizing the flow of road traffic
Is there a solution to the frequent traffic jams in modern metropolitan cities in the emergent behaviour of fire ants. Ant colonies intuitively maintain optimal flow without clogging passageways. Applying this understanding, along with modern technology and autonomous vehicles, could potentially solve urban traffic problems.
This article was first published in The Mint. You can read the original at this link.
One of the benefits (if you can call it that) of living in a rapidly exploding metropolitan city is that I have plenty of opportunity to observe how traffic jams develop. Many are the days that I have looked out of my office window and noticed that cars that, until moments ago were flowing smoothly, have begun to imperceptibly slow down until they eventually crawl to a stop, unable to move any further because of the sheer number of vehicles on the road.
From that point on, the jam gets steadily worse as autorickshaws and two wheelers insinuate themselves into the interstices between larger vehicles in a misbegotten attempt to get ahead—till the entire road becomes one immovable, gridlocked mass of metal.
I have often wondered why we haven’t yet cured this urban malaise. After all, the productivity costs of having commuters stuck in interminable traffic jams must surely be a problem worth solving.
A recent article in the journal Science, that describes how fire ants navigate through the tunnels and passageways of their anthills, offers some insights into a possible solution. However, to do that we need to find a way to apply the emergent behaviour of ants to our highways.
Ant colonies intuitively work to achieve the optimal rate of flow that will allow them to operate at the highest possible efficiency while, at the same time, ensuring that they do not clog up the narrow passageways of their home and bring the flow of traffic to a standstill. That they are able to achieve this with no central coordination—without a traffic policeman supervising the intersections or regulating the flow of traffic—is one of the more remarkable examples of the emergent properties of swarms.
The article refers to a study of a colony of 150 ants in which it was observed that, at any one time, no more than 10-15 ants were at work. This apparent tardiness of the majority of the colony had to do with achieving optimal flow so that the entire colony operates at maximum efficiency.
Once the number of ants needed to optimally achieve the required output without causing a clog in the flow of traffic had been reached, they seemed to know that if even one more ant entered the tunnel, the entire system would eventually seize up. After that point it was observed that ants would rather turn back than stay in the tunnel and clog up the system.
This instinctive understanding of optimal flow—even if it meant keeping the majority of the swarm idle—is what keeps ant colonies functioning at peak efficiency.
If we are able to ensure that our road systems are never loaded with more than the precise volume of vehicles that the system can sustain before it starts to seize up, we might have found the answer to our traffic problem. To do that, we will need accurate knowledge of the point beyond which even one more car getting on to the road would cause traffic to slow down and eventually stop—and then design our traffic systems to keep vehicles off the road once we reach that point.
Traffic engineers call this the fundamental diagram of traffic flow, a macroscopic model that balances traffic flux (the number of vehicles per hour) with the traffic density (the number of vehicles per kilometre) to determine what the optimal flow through the entire system should be.
Up until recently, it has been very difficult to calculate what the values of flux and density are in real time. Without live data there is no practical way in which to tell drivers that though they are getting on to a road on which traffic is plainly moving smoothly, by getting on they will be responsible for slowing it down and ultimately creating a jam.
Modern technology is increasingly changing this. Map software today provides us with accurate information about how bad the traffic is along our chosen route of travel, so much so that our digital assistants often alert us that we have to leave a bit earlier than planned for our meetings when traffic begins to build up en route.
After we head out, these applications are constantly evaluating our current speed, the congestion up ahead and the distance to our destination, re-calibrating the time at which we are likely to arrive and offering, on the fly, alternate routes that will get us where we need to go faster.
As useful as all this is, these technologies merely report on conditions as they develop and cannot pre-emptively direct traffic away from road systems that are approaching their critical point.
As our vehicles become more connected, we will be able to model the fundamental diagram of traffic flow to the point where it will become possible, at an individual level, to inform specific vehicles that if they insert themselves into a smoothly flowing traffic system, they will disrupt the delicate balance between flux and density and bring the entire system down.
Emergent design will then direct cars onto tertiary road systems instead of the highway, keeping them from disrupting the smooth flow of traffic until at some point in the future, the flow eases up and allows them to rejoin.
It is easy to imagine how much better things will become once we make the transition to autonomous transportation. Self-driving vehicles can communicate with each other in real time and will be able to autonomously compute flux and density in the same emergent way that ants do.
Then, dare I say, we might see the end of traffic jams.
