No abstract

This paper describes the design of an active gravity balanced planar mechanism, where auxiliary parallelograms are used to physically locate the center of mass of the mechanism. A sliding carriage positions a counterweight directly above the center of mass of the mechanism in order to make the system gravity balanced. The sliding carriage uses joint encoder data from the mechanism to compute the location of the center of mass. Experiment results demonstrate the effectiveness of the approach.

No abstract

No abstract

A reactive N-step look-ahead scheme is used to plan paths for an unmanned air-vehicle (UAV) to maximize the probability of successfully identifying an object on the surface. UAVs are used for reconnaissance, search and rescue, and surveillance missions where the quality of camera imagery can determine the success or failure of an operation. Reactive path planning and mathematical models for calculating the Probability of Detection, Recognition, and Identification (P-DRI) have been studied independently, but never merged to plan paths for UAVs. Combining these techniques results in the highest probability of successful object identification from aerial video. We utilize the Targeting Task Performance (TTP) metric to estimate the P-DRI and successfully incorporate it into a look-ahead path planner. We present both simulation and experimental results.