

Motion Planning 

While driving a car, we do not have unambiguous rules or equations to follow; everyday driving is an inexact science.
However, a carlike robot that drives itself must have exact motionplanning algorithms. We have to give a distinct algorithm for each given situation and requirements. MotionLab has succeeded in it, as you can see video demonstrations in this website. For instance, the motionplanning algorithms of Maze and Parallel Parking are totally different.
Furthermore, MotionLab has discovered a complete mathematical model on designing motionplanning algorithms; ALL the algorithms can be composed of only four kinds of atomic motioncontrol methods.
 We have discovered that any carlike vehicle’s motions can be composed of the following four motioncontrol methods: (1) CurvatureBased Method, (2) DirectionBased Method, (3) LineBased Method, and (4) OmegaBased Method.
 All the methods are independent each other: each method cannot emulate any another.
 A motion may be composed of a multiple number of these atomic methods.
The new math, Symmetric Geometry, has led us to this result. The use of the motioncontrol methods embody precise, safe, stable, and gentle Swan motions, as you can see. This mathematical model is hardware independent.






