Understanding the Science of Running
Running is not as simple a process as it seems at first glance. The physics of this type of movement is very different to the physics of walking, and as a result the leg and body movements are radically different. If we wish to recreate realistic running in either embodied robotic form, or via sequences for VR, then we need to understand this process intimately.
The 2012 Olympics have been a boon in this area. With so many athletes competing, and the increase in our general technological understanding in the four years since the last one, a good deal of research has been focussed on how we run. The news buzz behind the South African double amputee who ran in the olympics, and had planned to for at least five years, greatly helped pique interest in this kind of research, as an unintended beneficial side-effect.
Enter Young-Hui Chang, an associate professor of Georgia Tech, and who runs the universities running lab, officially called the Comparative Neuromechanics Laboratory. He has been studying the effects of running on the body, and discovered several interesting elements. Primarilly he has found that running is very similar to a bouncing ball. As he puts it: When humans, horses and even cockroaches run, their center of mass bounces just like a pogo stick.
This bouncing effect is caused by a very different combination of muscle movements when compared to walking, or even jogging. When you run, especially if you run all out, your muscles align in tune with one another so that the hip, knee and ankle joints all flex and extend at the same time when the foot hits the ground. Many of the leg muscles are turned on simultaneously, creating force and propelling the runner into the air.
The greater the force, the greater the speed, said Chang. Sprinters and coaches are constantly studying ways to move leg muscles and joints as quickly as possible so that a runner can hit the ground as hard as possible.
Elite runners and weekend joggers are able to consistently land with the same force, step after step. However, Changs research reveals that a stride is just like a fingerprint: no two are exactly alike. The torque generated by each joint is never the same. As a result, your legs have a mind of their own.
Your knee, for example, automatically adjusts its own torque, each step, based on what the ankle and hip do, said Chang. All of this happens without your brain getting directly involved. Your joints talk to each other, allowing you to concentrate on other things, like having a conversation or watching for cars.
For those with a background in VR or robotic control, this may sound a lot like inverse kinematics. That is of no surprise, for that is precisely what is going on. Inverse kinematics takes the goal position as the desired state, and works on moving every relevant joint into that position, thereby creating the desired state – rather than moving the joints and seeing what the final position is afterwards. This work strongly suggests that inverse kinematics should be the default state for computing running motions in avatars.
Obviously this research most directly impacts of prosthetic limb creation, which is exactly what Chang and his team plan to apply it to. The implications for VR avatars are however, no less obvious.
Further, by reverse engineering their study of running, particularly how the individual nuances flow, the team hope to improve our understanding of walking particularly the different gaits that different people have, and why they form in the first place. This is also something of great benefit to avatar movement and locomotion as well as prosthetic design.
It may seem backwards to fully understand the nuances of running before we study walking, but walking mechanics are actually more complex. Different muscles are activated at different times in a gait cycle. Joints dont move in unison. There is no bouncing ball phenomenon for walkers, Chang said.