THE USE OF THE WORD ROBOTIC IMPLIES A RIGID BULKY AND COMPLEX DEVICE. HOWEVER WE BELIEVE THAT A CONFLUENCE OF FACTORS NOW MAKES IT POSSIBLE TO SHED THIS STEREOTYPE AND BUILD ROBOTS THAT EXPLOIT THE PROPERTIES OF SOFT DEFORMABLE MATERIALS. THE GOAL OF THE PROPOSED WORK IS TO ENABLE A CLASS OF SOFT ROBOTS WHERE ALL THE FUNCTIONAL ELEMENTS ARE EMBEDDED IN A REMOVABLE AND TRANSFERABLE SKIN. WE ENVISION ACTIVE ELASTIC SKINS WHERE THE 2D ACTIVE ELASTIC SKIN MAY BE WRAPPED AROUND ANY HIGHLY DEFORMABLE 3D OBJECT TO CREATE A HIGHLY FUNCTIONAL SOFT ROBOT. IN THIS VISION ANY DEFORMABLE MATERIAL MAY BE THE FOUNDATION OF A SOFT ROBOT SUCH AS AN INFLATABLE BALLOON OR BLOCK OF FOAM. ACTIVE ELASTIC SKINS WILL BE FABRICATED IN 2D TRANSPORTED IN 2D AND THEN POSITIONED ON AN INERT 3D HOST TO ENABLE A FULLY FUNCTIONAL 3D ROBOT. ACTIVE ELASTIC SKINS WILL ENABLE MORE EFFICIENT TRANSPORTATION OF ROBOTS INCREASED VERSATILITY IN ROBOTS AND NEW PARADIGMS IN SOFT ROBOTIC DESIGN. FOUR RESEARCH OBJECTIVES ARE PROPOSED IN THIS PROJECT: (1) DESIGN AND FABRICATION OF ACTIVE ELASTIC SKINS WITH CONTROLLED BIAXIAL STRESSES: OUR AIM IS TO CREATE ACTIVE ELASTIC SKINS THAT CAN PRODUCE ENOUGH FORCE ALONG A GIVEN AXIS TO IMPART MOTION ONTO A DEFORMABLE HOST. AS AN INITIAL TARGET WE WILL STRIVE FOR SYSTEMS THAT PRODUCE BIAXIAL DEFORMATIONS OF AT LEAST 50% WITHOUT LOADING. WE BELIEVE THAT NO SINGLE ACTIVE MATERIAL WILL MEET THE ROBUST REQUIREMENTS OF SOFT ROBOTIC MOBILITY AND MANIPULATION TASKS. RATHER AN APPROPRIATE ACTUATION SYSTEM WILL BE MADE FROM COMPOSITE ACTIVE MATERIALS THAT USE PARALLEL RESPONSES TO CONTROL MOTIONS. IN THIS PROJECT WE WILL USE SHAPE MEMORY ALLOYS AND SHAPE MEMORY POLYMERS AND LEVERAGE THEIR INTERACTION WITH ELASTIC SUBSTRATES TO ACHIEVE HIGH-DEFORMATIONS. (2) SYNTHESIS CHARACTERIZATION AND FUSION OF ELASTIC SENSORS: HIGHLY DEFORMABLE ROBOTIC SYSTEMS REQUIRE EQUALLY DEFORMABLE SENSORS FOR STATE ESTIMATION AND TACTILE FEEDBACK. WE WILL LEVERAGE OUR ONGOING WORK ON SOFT STRAIN PRESSURE AND CURVATURE SENSORS TO CREATE MULTIMODAL AND MULTILAYERED PROPRIOCEPTIVE SKINS. THE UNIQUE FEATURES REQUIRED IN THE PROPOSED SYSTEM ARE HIGH DEFORMATIONS (>50%) AND MULTIPLE SENSOR ELEMENTS INTEGRATED INTO A SINGLE SUBSTRATE. WE WILL EMPLOY OUR EXISTING SINGLE-SENSOR PROTOTYPES TO DESIGN AND PRODUCE FILM-LIKE MULTIMODAL CURVATURE- AND TACTILE-SENSITIVE SHEETS. THIS WILL REQUIRE CHARACTERIZATION OF SENSOR INTERACTIONS AND PROCESSING TECHNIQUES FOR ELASTOMERS AND LIQUID METALS FOR HIGH-DENSITY SENSOR ARRAYS. (3) DEVELOPMENT OF EFFICIENT ALGORITHMS FOR STATE ESTIMATION: SOFT ROBOTS PRESENT A UNIQUE CHALLENGE BECAUSE OF THE HIGH NUMBER OF DEGREES OF FREEDOM COMPARED TO CONVENTIONAL SYSTEMS. TO ADDRESS THIS POINT WE WILL BUILD OFF OF OUR EXPERTISE IN COMPUTATIONAL GEOMETRY SIMULATION OF HIGH DEGREES OF FREEDOM AND COMPLIANT CONTROL TO DEVELOP MORE EFFICIENT ALGORITHMS CAPABLE OF TRANSFORMING LARGE ARRAYS OF SENSOR INPUTS INTO AN ESTIMATION OF SYSTEM STATE USING EMBEDDED HARDWARE. ONE APPROACH WILL BE TO EMPLOY DENSE ARRAYS OF SENSORS SO AS TO OVER-INFORM THE CONTROLLER MAKING INTERPOLATION OF STATE BETWEEN DISCRETE SENSORS LESS COMPUTATIONALLY TAXING. (4) INTEGRATION AND DEMONSTRATION OF SOFT ROBOTIC LOCOMOTION USING ACTIVE ELASTIC SKINS: BY INTEGRATING THE ACTUATION SENSING AND CONTROL INTO A SINGLE ELASTIC SHEET WE WILL DEMONSTRATE TWO LOCOMOTION STRATEGIES USING THE SAME ACTIVE ELASTIC SKIN POSITIONED AROUND AN INFLATED CYLINDRICAL BALLOON: PERISTALTIC LOCOMOTION AND INCHWORM LOCOMOTION. FURTHERMORE WE WILL SHOW THAT THE ACTIVE ELASTIC SKIN MAY BE REMOVED AND THE BALLOON DEFLATED SUCH THAT THE SOFT ROBOTIC SYSTEM TAKES UP MINIMAL SPACE REQUIREMENTS DURING TRANSPORT.