SML > Research People > Alan Wilson > CHEETAH robot
Bio-Inspired Robots
Robot designers look to nature for inspiration for the design of dynamic, free ranging and legged robots. One company in particular, Boston Dynamics, is leading the field with their range of innovative robots, including CHEETAH, the world’s fastest legged robot.
Alan Wilson’s role in the CHEETAH project is to discover and translate the mechanics of cheetah locomotion into engineering principles that can be used by robot designers. By understanding the basic principles of how animals run, remain stable and use their muscles it should be possible to make legged robots that are faster, more capable on varied terrain and more economical.
The scientific goals of this work are to:
Explore the dynamics of gallop locomotion, particularly the ground reaction forces and the fluctuations in mechanical energy that result. Alan’s team are interested in the distribution of limb contact timings and the forces applied by each limb to elucidate how legs are combined to deliver smooth motion of the centre of mass with minimal pitching and rolling movements. This is undertaken during steady speed galloping, acceleration and manoeuvring.
These data are used to inform and validate modelling by colleagues at Boston Dynamics and to demonstrate the biological solution to high speed manoeuvring locomotion. Video and a report on early cheetah measurements at the Zoological Society of London Whipsnade Zoo are available at http://news.bbc.co.uk/1/hi/sci/tech/8137962.stm.
Explore the inherent stability of locomotion by determining the body motions during many strides and at high speed. For this work the team use a 30 gram GPS inertial measurement unit (IMU)(10 Hz raw data GPS and accelerometers, gyroscopes and magnetometers) attached to a harness or collar on the cheetah. (figure 1). The data from the unit give measures of speed, position, acceleration and orientation.
From these measurements angular velocity and centripetal acceleration (more robust experimental measures than velocity and radius) can be calculated during the many twists and turns that the cheetah perform when hunting. During such turns, the additional centripetal accelerations that the animals experience result in them effectively running in an increased gravity environment, increasing the peak forces their limbs must withstand and potentially limiting gallop speed.
In wild cheetah these data will be used to examine the relationship between subsequent strides to explore how natural perturbations due to surface topology, speed change and turning are handled so that smooth locomotion is maintained. Such data, whilst less fine grained than force plate recordings, are valuable in validating both reductionist and more complex models of high speed and unsteady locomotion.