The flailing of a gymnast who's missed a step on the balance beam might not be far off from what the rest of us experience every day. Each step we take is really a tiny fall, a mathematical model suggests. The random-looking variation in our footfalls is actually a series of corrections. Our strides are all screw-ups—but thanks to the fixes that happen without us knowing, our walking routines look like a perfect ten. Manoj Srinivasan, who runs the Movement Lab at Ohio State University, and Yang Wang, a doctoral student at the time, studied walking down to the millimeter. They put motion-capture markers on people's feet and pelvises, as if preparing their legs to star in Avatar 3. Then the 10 subjects walked on treadmills at various speeds, while cameras captured every motion. The data showed that no one was a perfect walker. "Every step is slightly different from every other step," Wang says. Rather than placing their feet in the same spots on the treadmill with each stride, subjects constantly deviated to the left or right, front or back. Wang and Srinivasan set out to create a model that would explain this variation. Is it as random as it looks, or are people's footfalls predictable? "We were trying to find a mathematical relation between the next foot position and the current pelvis state," Wang says. Tiny leftward or rightward shifts in the pelvis represent the shifting of that whole heavy torso we have to balance on top of our legs. When the researchers compared these pelvis movement to variations in footfalls, what they found wasn't random at all. More than 80% of the side-to-side variation in foot placement could be predicted by movements in the pelvis beforehand. When the top half of the body starts to tip left, we set our next step a little bit farther left to push ourselves back. When we wobble right, we step wider in that direction. With each stride, we start to fall, and "we are constantly making little corrections to be stable," Wang says. The researchers even found that tracking the pelvis predicted where a person's feet would fall better than tracking their actual feet. Halfway through each stride, they write, "the pelvis ‘knows’ much more about the future foot position than the foot itself." They ruled out other possible influences too, like where a person was on the treadmill belt. The single best explanation for a person's pattern of footfalls was self-correction. Understanding how humans walk could help engineers develop better human exoskeletons and walking robots, Wang and Srinivasan say. And knowing how we keep ourselves stable might someday help us prevent falls in people who are vulnerable to them, such as the elderly or those with movement disorders. There's always self-doubt to trip us up, though. Now that they know they're falling all the time, do the researchers doubt their own bipedal abilities? "Honestly, not so much," Srinivasan says. "These series of errors and recoveries are pretty sub-conscious for a healthy adult." However, he speculates that to stay safe, maybe "you can't think about it too much."
Image: Richard Leeming (via Flickr)
Wang, Y., & Srinivasan, M. (2014). Stepping in the direction of the fall: the next foot placement can be predicted from current upper body state in steady-state walking Biology Letters, 10 (9), 20140405-20140405 DOI: 10.1098/rsbl.2014.0405