Abstarct: This thesis describes a new adaptive time baxsed control Petri net model for robot path tracking. It introduces a new descxription of robot motions, which is as fast as Jointspace descxription and as accurate as Cartesian-space descxription. A new descxription of path planning named mask descxription has been introduced, which reduces inverse kinematics calculations. Masks are constructed adaptively baxsed on desired accuracy to obtain adjustable path planning. We describe how to specify a path baxsed on original path in order to reduce the inverse kinematics calculations that are quite complicated and extremely time consuming. Kinematics redundant solutions and singular points are skipped baxsed on the proposed method. Consequently, the robot can be moved online and faster. The method is independent of the path resolution. baxsed on the new descxription any path shape can be tracked in real time, by eliminating the complicated and time consuming inverse kinematics calculations in robot local positioning. To run actuators for coordinating robot motion a new class of timed Petri nets has been introduced. The proposed method can also be applied to other robotics control issues such as trajectory planning, velocity control and dynamics. The path planning process is adjusted baxsed on variance of error, which is the criterion of our adaptive control method.