首页 >> 小红猪作品 >> 文章



Sometimes life runs more smoothly when you stop trying to control it. Mark Buchanan goes with the flow

Why complex systems do better without us

 ●WE HUMANS prefer the tidy to the untidy, the ordered to the disordered. We like pristine geometrical regularity, and eschew what is erratic and irregular. We want predictability and, more than anything, we want control.

In these confusing times, it might seem as if we have little power over anything. Instead of letting it get us down, though, perhaps we should take comfort from the work of Dirk Helbing , a physicist at the Swiss Federal Institute of Technology (ETH) in Zurich. Helbing has been studying the movement of tens of thousands of cars on road networks; the workings of vast webs of interacting machines on factory floors; and other systems, where the complexity of what happens and why routinely defeats the human mind.

What Helbing and others are finding is that our penchant for regularity and control is seriously misguided. In many situations they are discovering that it is better to give up some of our control and let systems find their own solutions. Often the answers turn out to be unlike anything our minds would imagine, yet the outcomes are far more efficient.

The findings come as something of a relief to today’s engineers, who are increasingly dealing with problems too complicated for them to solve. Take one of the earliest successes chalked up by machines allowed to take control.

Back in 1992, General Motors were having trouble managing the automated painting of trucks at an assembly plant in Fort Wayne, Indiana. Machines in 10 different paint booths could paint trucks as they came off the line, but because the trucks came off in an unpredictable order and the painting machines needed sporadic maintenance and repair, finding an efficient assignment of trucks to booths seemed impossible.

General Motors’ visionary engineer Dick Morley suggested letting the painting machines find a schedule themselves. He set out some simple rules by which the various machines would “bid” for newly available paint jobs, trying their best to stay busy while taking account of the need for maintenance and so on. The results were remarkable, if a little weird. The system saved General Motors more than $1 million each year in paint alone. Yet the line ran to a schedule that no one could predict, made up on the fly by the machines themselves as they responded to emerging needs.

Production processes generally depend on so many inputs, parameters and factors that even small changes in the set-up can lead to wildly different and unpredictable consequences. That is why it is almost impossible to predict what will happen in a new production line based on previous experience. “Managers sometimes take performance in past set-ups and try to estimate what will happen in a new setting by interpolation,” says Helbing. “This often gives very bad results.”

To cope, he says, engineers need a healthy respect for the complex unpredictability of these systems and how natural human inclinations often lead to undesirable outcomes. “You can’t steer these things like you can a bus,” says Helbing. “You have to learn to use the system’s own self-organising tendencies to your advantage.”

Helbing has come to this view by an unusual path. Though he trained as a physicist, he became fascinated in the early 1990s by parallels between physics and human movements. “I was inspired by the similarity between fluid flows and how people walk around obstacles,” he recalls. For nearly two decades, he and colleagues have been studying the mathematics of collective human motion, which explains why Helbing now holds a chair in sociology.

Social scientists usually focus on the variability of human behaviour, which is hard to predict. But Helbing argues that in many cases it isn’t very important. That’s because circumstances often constrain peoples’ options so much that humans respond almost automatically to external forces, making their average behaviour predictable. On the roads, for instance, people generally drive close to or just over the speed limit, similar to the way self-propelled particles repel one another when they get too close.

Although the behaviour of individuals is often simple, the collective patterns to which it leads can be counter-intuitive, making common sense a faulty guide to what might happen. For example, it is generally true that traffic jams become more likely as traffic density increases. It’s not always the case, though, as Helbing’s group has shown.

Consider a two-lane road carrying both cars and trucks, where the cars are moving faster on average. At low traffic densities, the cars have plenty of space to overtake and can easily pass the trucks. As the traffic density increases, drivers find it more difficult to overtake because other vehicles are in the way. However, evidence from simulations and real traffic flows shows that at a critical density of traffic, the obstruction to lane-changing begins to have a beneficial effect. Because drivers tend to stay in one lane, they disturb the flow of traffic less, leading to a higher total throughput of vehicles.

Similar counter-intuitive results show up in crowds of people. In simulations and experiments, Helbing’s team has confirmed what they call the “slower-is-faster” effect. When people try to escape from a room through a doorway, more get out if everyone stops rushing as this prevents obstructions. Surprisingly, it turns out that placing an obstacle in front of the door can actually enable people to get out faster, as it helps to regulate the flow of people and maintain its fluidity. “A suitable obstacle can improve the outflow by about 30 to 40 per cent,” says Helbing.

What makes it work is that crowds adjust to local conditions. When two streams of people meet at either end of a narrow passage, you might expect a jam to form as only a chaotic trickle of people pass through. But in real life, people often do something completely different: they organise so that a group goes through first in one direction and then the other, as long as the density isn’t too high. The crowd organises itself spontaneously to a better outcome.

Helbing has found that you can model crowds using ideas akin to those from physics. As a queue grows on one side of the passage, it produces something resembling the pressure of a fluid or a gas. A high density of people pressing together ultimately acts to drive people through the opening, thereby relieving the pressure.

Further work has convinced him that systems involving pedestrians, traffic and products flowing through factories often work in surprisingly similar ways, hence lessons learned about one may also apply to another.

Last year, Helbing and Stefan Lammer at the Technical University of Dresden in Germany began wondering if traffic lights could also be engineered to cut congestion. According to a report by David Shrank and Tim Lomax at the Texas A & M University in College Station, congestion in the US alone costs an estimated $78.2 billion, wastes 4.2 billion hours in delays and 10.9 billion litres of fuel. So the potential impact of efficient traffic flow could be huge.

This would mean giving traffic lights a way to adapt their behaviour, which most of today’s systems lack. At the moment, engineers force traffic into patterns that appear favourable. Lights on main roads stay green longer during peak hours, for example. But it’s the engineers who do this tuning based on average conditions observed in the past; most traffic lights don’t have the flexibility to respond to changing conditions on their own. Engineers also take some things for granted, such as the notion that lights must be managed from a central control.

Lights can do a better job, Helbing and Lammer have found, if they are given some simple operating rules and left to organize their own solution. To demonstrate this, they developed a mathematical model that assumed traffic flowed like a fluid, a wellestablished traffic engineering technique. The model also describes what happens at road intersections, where traffic entering from one road has to leave by another, much like fluid moving through a network of pipes.

Of course, jams can arise if traffic entering a road overloads its capacity. To avoid this, Helbing and Lammer make the lights at each intersection respond to growing traffic pressure, like the people going through the passage. Each set of lights carries sensors that feed information about the traffic conditions at a given moment into a computer, which then calculates the flow of vehicles expected in the near future. The computer also works out how long the lights should stay green in order to clear the road and relieve the pressure. In this way, each set of lights can estimate for itself how best to adapt to the conditions expected at the next moment.


Best left alone

This isn’t enough, however, because the lights might adapt too much. If they are only adapting to conditions locally, they might cause trouble further away. To avoid this, Helbing and Lammer have devised a scheme whereby neighbouring lights share their information so that what happens around one traffic light can affect how others respond. By doing so, the self-organised lights prevent long jams from forming.

Despite the simplicity of these rules, they seem to work remarkably well. Helbing and Lammer have demonstrated in simulations that lights operating this way should achieve a significant reduction in overall travel times and keep no one waiting at a light too long (See diagram, below等你们抢到了我再给你看图吧……不会贴。). Nonetheless, the behaviour of the lights doesn’t generally fit with human notions of what ought to be efficient. “How long lights stay green is unpredictable,” says Lammer. Yet the average journey times go down and become more predictable.

What’s more, the scheme eliminates other irritating problems that afflict traditional traffic control. At quiet times, drivers typically have to wait far longer than is really necessary at intersections because the lights’ schedules are designed to serve a large number of vehicles. And in the middle of the night, lights keep stopping cars even when there is no need. The self-organising traffic scheme eliminates these problems because the lights remain responsive to local demands, for instance sensing an approaching car and changing to green to let it through.

Town planners are beginning to look at self-organising lights as a practical solution to looming traffic congestion. Helbing and Lammer are working with a local traffic agency in Dresden, Germany, first to test and then hopefully to implement the idea. In early simulations based on Dresden’s road layout, they have had encouraging results. “We’ve found significant reductions in waiting times and fuel consumption, and we can also accelerate public transport,” says Lammer. Authorities in Zurich, Switzerland, have also been taken by the idea.

Yet Helbing and Lammer suggest their scheme only begins to illustrate the potential for self-organised traffic flow. The technology to make cars also sense and respond to local conditions already exists, and many of us may soon cede at least some control of our car to on-board guidance systems. If cars can talk to one another, Helbing and his colleagues have shown, they could improve traffic conditions even more – greatly reducing the severity of jams at times and possibly even eliminating them altogether (see “Cruise Control”, below).

The wider lesson is that we just can’t trust our intuition when it comes to the supercomplex systems that we depend on today. We may never learn exactly how to control these systems in the traditional fashion and the best way to cope may be by learning new principles for letting them manage themselves. Engineering isn’t just about solving problems any more, but building systems that can solve their own problems. Being in control, it seems, may increasingly demand being a little out of control.

 ●Mark Buchanan’s latest book is The Social Atom (Cyan Books, 2007)

Further Reading: Self-control of traffic lights and vehicle flows in urban road networks by Stefan Lmer and Dirk Helbing, www.arxiv.org/abs/0802.0403


Cruise Control

Some of today’s cars already contain technology that lets a driver hand over some control to on-board devices. Unlike conventional cruise control, which simply maintains a driver’s chosen speed, “adaptive” cruise control (ACC) uses radar to sense the distance and speed of the car in front. By updating that information several times a second, the system maintains the car’s speed and separation distance and automatically brakes if the car in front slows down, or accelerates when the leading vehicle does. Also, it responds faster and more accurately than human reflexes.

In recent simulations, engineer Arne Kesting at the Technical University of Dresden in Germany, working with Dirk Helbing at the Swiss Federal Institute for Technology in Zurich and others, have studied how the technology might help traffic respond to emerging problems. While it may be some time before most automobiles on the highway are fitted with adaptive cruise control, the researchers have shown that even a small fraction of users could make a huge difference.

At the moment, drivers can respond only to conditions they encounter or perhaps hear about on radio reports. ACC cars could easily be fitted with sensors able to receive signals conveying local traffic conditions from roadside monitors or other cars. Kesting and his colleagues suggest that these signals could reduce congestion by making cars equipped with ACC drive more intelligently.

For instance, cars flowing out of a traffic jam could automatically drive closer together in order to clear the jam faster. Meanwhile, cars approaching the jam would slow more gradually, rather than brake abruptly when reaching it. This would maintain greater fluidity in the traffic, improving road capacity and stability of traffic flow. Kesting’s simulations suggest that if 25 per cent of the cars were using ACC, the scheme could eliminate many traffic jams. Even if only 3 per cent of the cars were equipped, travel times could be significantly reduced.

Kesting and Helbing are currently testing these ideas with Volkswagen and hope to see the scheme on real roads in a few years.


9 Responses to “Why complex systems do better without us(《新科学家》8月6日封面文章)【小红猪翻译小分队】”

  1. [...] 《Why complex systems do better without us》,《新科学家》8月11日封面文章。公路上的科学:有时候你少管一点,结果会更好…… [...]

  2. fwjmath说道:


  3. 桔子说道:


    • fwjmath说道:


    • fwjmath说道:


    • fwjmath说道:


  4. [...] 原文请看这里作者介绍:fwjmath,只会学术的宅男×1,正业为数学,爱好捣鼓自然科学,在外求学中,算是大二。个人博客是http://fwjmath.spaces.live.com。感谢提出修改意见的QueenKerene同学和桔子同学。 [...]

  5. sunsagittary说道:


  6. emergence说道:

    你好!不知Why complex systems do better without us这篇文章翻译了没有,如果没有,我愿意翻译一下。


Leave a Reply