Swing Shift: A Bridge Wobbled and Robotics Benefited

Commuters cross the Millennium Bridge in London. 

Commuters cross the Millennium Bridge in London. 


Anthony Devlin/Zuma Press

When London’s Millennium Footbridge opened 18 years ago, the 1,066-foot span swayed so badly it was closed almost immediately.

Ever since, engineers who study human locomotion have puzzled over a conundrum: Why did pedestrians who were tripped up by the trembling bridge walk in a way that made the wobble worse?

The answer has helped researchers design robotic prostheses and wearable exoskeletons to improve the mobility of stroke victims, amputees and others.

“It’s a road map,” said Gregory Sawicki, a mechanical engineer at Georgia Tech who designs exoskeletons to support the weak ankles of the elderly. “We learn how humans do it first, and then we either steal or improve on the best ideas.”

Take a Walk

The energy cost of walking goes up when 400 pedestrians shake a bridge, compared with 80 who don’t, but the cost lessens as the pedestrians synchronize their steps with the swaying of the bridge.

Metabolic energy cost per unit distance



400 pedestrians


80 pedestrians









Source: Manoj Srinivasan and Varun Joshi, Ohio State University

Around 80,000 people streamed across the Millennium Bridge when it opened, with as many as 2,000 on it at once. When it started to shimmy, the pedestrians tended to widen their stances, walk in sync with each other and time their steps to the rocking of the structure.

The bridge was expected to sway a bit, but it ended up shifting in each direction by as much as 3 or 4 inches.

(The bridge reopened in 2002 after the addition of shock-absorbing dampers.)

“The classic strategy is placing your foot roughly in the direction in which you are falling,” said Manoj Srinivasan, a mechanical engineer at Ohio State University who has studied how pedestrians responded to the bridge. “If you’re falling sideways, you put your foot to the side.”

The pedestrians’ awkward gait kept them upright. But why did they stick with it if it also amplified the bridge’s movement?

According to Dr. Srinivasan, humans deploy two strategies when they walk and run: They try to avoid falling down, and they try to conserve energy.

Walking in synchrony with lateral bridge oscillations, he said, lowers the energy cost.

In other words, synchronizing footsteps with the swaying bridge may have increased its rocking, but with enough people, it also may have helped them conserve energy.

“There is tons of evidence that people prefer to walk in ways that minimize the amount of food needed to accomplish the task,” said Max Donelan, a neuromechanics professor at Simon Fraser University in Vancouver.

In their latest study, published last week in the journal Biology Letters, Dr. Srinivasan and Varun Joshi, a researcher in his lab, built upon their previous research by modeling how the pedestrians adjusted their stride in response to feedback from the environment.

We learn how humans do it first, and then we either steal or improve on the best ideas.

—Gregory Sawicki, a mechanical engineer

“This goes on all the time,” Dr. Donelan said. “If you reach into the fridge to pick up a jug of milk that somebody has taken a cup out of since you last grabbed it, it’s lighter than you remember. You realize you pulled it too hard. You had expectations about the weight of the jug of milk.”

Unraveling how people combine the desire for stability, the urge to conserve energy and feedback from a bridge’s sway or a milk carton’s heft helps engineers design better devices.

“If we thought the goal was to be stable all the time, we might build devices that would compete against their actual desire,” Dr. Donelan said.

Interest in this area of research, which is relatively new, was spurred in part by the military.

Not long after London’s shuddering footbridge shook the confidence of its pedestrians, the U.S. military, coincidentally, started investing in exoskeletons to help soldiers carry their loads.

“That turned the dial up,” Dr. Sawicki said.

Because the exoskeletons were cumbersome and used too much energy, the military lost interest—but the technology survived.

“That thing that cost soldiers a lot of energy was repurposed to help people who can’t walk in the first place,” Dr. Sawicki said. “It morphed into rehabilitation devices where energy cost isn’t the priority.”

So far, the products aren’t widely used by the general public, but some companies are marketing the devices. Among them are

Ekso Bionics

which makes exoskeletons for stroke and spinal-cord injury rehabilitation and for jobs that involve heavy lifting, and Seismic, which integrates robotics into clothing designed to improve the wearer’s strength and stability.

Eventually, engineers like Dr. Sawicki hope this kind of technology will rival or even surpass human capabilities.

For now, that’s a bridge too far.

“We’re not close to what the human body can do,” Dr. Sawicki said, “but we’re approaching it.”

Write to Jo Craven McGinty at Jo.McGinty@wsj.com