How might one sense the direction of gravity in a mobile robot? The only method I've thought of so far is a pendulum on gimbals with two shaft encoders on the gimbal hinges, the whole mounted in an oil-filled chamber to damp out oscillations. Alternatively, inspired by the semicircular canals in the inner ear, one might use mercury flowing in a circular tube mounted in the vertical plane, but mercury is fairly nasty stuff to work with. Load cells in the feet or wheels would work while the robot is stationary, but when it's moving, it might not be possible to disentangle the gravitational signal from the forces due to the robot's own acions.
- posted
17 years ago