FOR THE PROBLEM OF ROBOTIC MANIPULATION WHEREIN A ROBOTIC MANIPULATOR INTERACTS WITH OBJECTS OR ITS ENVIRONMENT USING AN ENDEFFECTOR (GRIPPER) THERE HAVE BEEN ATTEMPTS TO QUANTIFY AND OPTIMIZE HOW GOOD A HOLD THE GRIPPER HAS ON AN OBJECT BEING MANIPULATED USING GRASP QUALITY METRICS. TRADITIONALLY GRASP QUALITY ANALYSIS ASSUMES A RIGID GRIPPER AND A RIGID OBJECT. HOWEVER TECHNOLOGIES CURRENTLY BEING INVESTIGATED BY NASA INCLUDING CLIMBING ROBOTS ROBOTIC TRANSPORT OF SURFACE ASSETS AND ROBOTIC GRAPPLING OF FREE FLOATING OBJECTS RELY ON COMPLIANT GRIPPERS AND GRIPPERS WHO PRIMARILY APPLY PULLING AND SHEAR LOADS. OF PARTICULAR INTEREST ARE MICROSPINE AND GECKO ADHESIVE BASED GRIPPERS CURRENTLY UNDER DEVELOPMENT. FOR SUCH GRIPPERS TRADITIONAL GRASP QUALITY ANALYSES DO NOT APPLY. THE FIRST OBJECTIVE OF THIS PROPOSAL IS TO DEVELOP A STIFFNESS BASED APPROACH TO ANALYZE GRASP QUALITY ACCOMPANIED BY THE THEORETICAL BACKGROUND TO EVALUATE ITS EFFECTIVENESS. IT IS COMMON TO USE MULTIPLE GRIPPERS IN PARALLEL TO PERFORM MANIPULATION TASKS. IN THE CASE OF CLIMBING ROBOT LOCOMOTION IT IS NECESSARY AS EACH LIMB/GRIPPER MUST GRASP THE SURFACE AND WORK TOGETHER TO MOVE ALONG ITS TRAJECTORY. TO ACCOMMODATE UNCERTAINTIES IT IS DESIRABLE TO FIND GRASP REGIONS RATHER THAN SINGLE GRASPS IN WHICH THE GRIPPERS CAN BE PLACED TO ACHIEVE THEIR TASKS. A NOVEL METHOD FOR FINDING SUCH REGIONS IS TO FRAME THE PROBLEM AS A BIPARTITE GRAPH PROBLEM WHERE GRAPH EDGES REPRESENT GRASP QUALITIES. FROM THIS BASIS FINDING GRASP REGIONS EQUATES TO FINDING BICLIQUES CORRESPONDING TO REGIONS OF DESIRABLE SURFACE GEOMETRY. THE NEXT OBJECTIVE OF THIS PROPOSAL IS TO DEVELOP ALGORITHMS FOR FINDING REGIONS THAT ARE PRACTICALLY USEFUL FOR GRASPING BY UTILIZING AND EXPANDING ON CURRENT GRAPHICAL AND MANIPULATION ALGORITHMS. PRACTICAL IMPLEMENTATIONS OF THESE ALGORITHMS MUST ALLOW FOR TIME-EFFICIENT COMPUTATION OF GRASPS FOR USE IN ROBOT PLANNING. THE OBJECTIVE IS TO USE TECHNIQUES INCLUDING PROBABILISTIC METHODS AND EXAMPLE-INFORMED SEARCHES TO EFFICIENTLY GENERATE GRASP REGIONS THAT MEET OR EXCEED A MINIMUM QUALITY THRESHOLD (SAFETY FACTOR) IN REAL TIME. FINALLY IT IS SOMETIMES NECESSARY TO PERFORM IN-SITU RE-GRASPS. FOR A CLIMBING ROBOT A GRASP MAY BEGIN TO SLIP OR PERFORM WORSE THAN EXPECTED AND RE-GRASPING OF THE SURFACE WILL BE NECESSARY. THE OBJECTIVE IS TO USE TOOLS SUCH AS IN-SITU PULL TESTING TO DETECT POOR GRASPS AND DEVELOP METHODS FOR USING THIS INFORMATION AS FEEDBACK TO EFFECTIVELY EXECUTE A RE-GRASP. DEVELOPMENT OF THIS TECHNOLOGY DIRECTLY IMPACTS NASA INTERESTS BY SUPPLYING METHODS OF ANALYZING AND PLANNING FOR ROBOT MANIPULATION AND CLIMBING ROBOT TASKS THEREBY ADVANCING CRUCIAL TECHNOLOGIES FOR ROBOTIC SPACE OPERATIONS AND EXPLORATION.