The most intellectually stimulated I got last week was on Wednesday night, over a sushi dinner with a friend. Naturally we were discussing the tragedy, and our conversation turned toward getting the airlines running again--she leads a bicoastal life, and air travel is a big part of her life.
With the news commenting on how difficult it would be to get air travel flowing again--planes were forced to land, the backlog of travelers needing to fly, heightened security measures, etc--we discussed the inefficiencies of travel even in a normal period, things like how bad weather in New York can cause rippling effects throughout the system, leading to delays around the country.
The problem seems to have two prongs that I know about--reliance on a relatively limited number of "hubs" through which all traffic is routed, and an exceedingly specific set of routes and plans, a "top-down" structure that crumbles when a single element is out of place.
Shortly before our conversation, I had finished Steven Johnson's new book Emergence: The Connected Lives of Ants, Brains, Cities, and Software, a humanist take on the science of complexity, illuminating how patterns of order--and, perhaps more importantly, success--emerge in similar fashion across a variety of complex systems.
That fashion, at it's simplest, is through massive numbers of brief local interactions that follow a few simple rules... Without explicit intention, these local interactions lead to often astounding order at ever-higher levels, usually to the effect of perpetuating that system... As Steven explains:
This local feedback may well prove to be the secret to the ant world's decentralized planning. Individual ants have no way of knowing how many foragers or nest-builders or trash collectors are on duty at any given time, but they can keep track of how many members of each group they've stumbled across in their daily travels. Based on that information--both the pheromone signal itself, and its frequency over time--they can adjust their own behavior accordingly. The colonies take a problem that human societies might solve with a command system (some kind of broadcast from mission control announcing that there are too many foragers) and instead solve it using statistical probabilities. Given enough ants moving randomly through a finite space, the colony will be able to make an accurate estimate of the overall need for foragers or nest-builders...
This book is a great primer on the subject of complexity, particularly powerful in how it uses subjects that many are familiar with (urban living, mass media, buying crap at Amazon), and explains how the processes in evidence there are the result of emergent systems.
Since my brain was metabolizing this book, I began thinking how such bottom-up thinking could inform airline systems. Complex adaptive systems are amazingly resilient--ant colonies, brains, and cities are remarkably stable, and when some trauma does occur, the system often simply routes around it and continues its development. Airlines, on the other hand, buckle under remarkably little stress, and become unmanageable under even moderate stress.
If airlines were to pattern themselves on emergent systems, they'd use smaller aircraft and a freer range of flight paths, such that the system can figure out the most efficient means of transporting people, instead of being restricted to the paths laid down "from above" by airline planners.
As it happens, I believe this is the thesis of James Fallows book, Free Flight. I don't know if Fallows is knowingly advocating an emergent system, or if he stumbled upon the notion independently.
Such a plan raises all kinds of questions, namely, "How do I know that I can get from San Francisco to New York at a certain time on a certain day?" We've all come to rely on schedules to make sure we get from point A to point B.
But what if, instead of us being beholden to the pre-determined whims of the air carriers, there was a system that would aggregate customer requests and turn those into flight plans? Such that I'd say, "I'd like to get from San Francisco to New York on March 15th, leaving at 9a." Actually, that's not interesting, 'cause those are two hub cities that get plenty of air traffic. But if I said, "I'd like to get from Santa Rosa, CA to Rochester, NY." And if enough other people made that request, it would achieve some kind of mass, and an entrepreneurial sort would offer to fill that request. Or figure out an as-optimal-route-as-possible that likely wouldn't be whatever a planner had decided.
Some interesting thoughts on complexity and airlines (specifically cargo routing) are contained in Manifest Destiny: Adaptive Cargo Routing at Southwest Airlines.
3 comments so far. Add a comment.
Previous entry: "Keeping in mind the human scale."
Next entry: "Which would you choose?"
very thought provoking piece peter - it reminded me of the way the internet works and one of the limitations that was overcome that makes it so viable - packet routing (i.e. splitting up data into chunks that get sent over the network through the most viable channels available). its unfortunate that routing people is not so easy (a la star trek's teleporter, perhaps?). i think the biggest problem (and one that perhaps deserves some thought) is one of scale - ants and packets, or even packages are not an exact science - losing a few here and there is not a big deal though as we all know, losing a plane (and its passengers) is.
Posted by evan @ 09/19/2001 10:32 AM PST [link to this comment]
I can't wait to read this book. On another note, this from a recent UXblog post via BBC:
Artificial Ants Solve Network Problems
Researchers have found that control programs based on the foraging behaviour of ants can keep data networks running more efficiently and cope with congestion better than many human alternatives.
Marco Dorigo and colleagues at the Free University of Brussels are creating small, smart computer programs based on ant foraging. Many individual ants may discover different routes to the same food but the shortest path that leads to it will have the strongest concentration of pheromone, a chemical indicator laid down by the ants. Professor Dorigo and his colleagues have created artificial ants that can lay and sniff virtual pheromone trails and can learn their way around a network. By simply laying and smelling the strength of the pheromones along each potential path, the ants swiftly generated maps that showed the fastest route to any end point.
When congestion was simulated on the artificial networks, the ants beat all the other popular routing systems in the speed with which they reconfigured the network to avoid the traffic jams. Already some companies are using the ant systems to do a better job of managing delivery networks and supply chains.
Posted by -challis @ 09/19/2001 12:54 PM PST [link to this comment]
Not sure how effective they are but perhaps car pooling/ride sharing sites like http://www.erideshare.com and http://www.rideshare.com perform a similar brokering service for the open road?
Posted by Angus Fraser @ 09/20/2001 12:17 AM PST [link to this comment]
Add A New Comment: