IN 2019, WITH A GLOBAL POPULATION OF 7.6 BILLION PEOPLE, THE WORLD FARMED OVER 690 MILLION HECTARES OF WHEAT, CORN, SOYBEAN, AND RICE. THE WORLD POPULATION IS PROJECTED TO RISE TO OVER 9 BILLION BY 2050, DRIVING INCREASING DEMAND FOR AGRICULTURAL PRODUCTS. SIMULTANEOUSLY, OUR IMPACT ON THE ENVIRONMENT AND A CHANGING CLIMATE BEG US TO REDUCE, OR AT LEAST HALT THE INCREASE IN LAND USED FOR FARMING. THE RESULT IS PRESSURE FROM ALL SIDES TO INCREASE THE EFFICIENCY OF AGRICULTURAL PRODUCTION. CONTINUOUS INNOVATION IN PLANT GENETICS, FARMING PRACTICES, AND ROBOTICS AND AUTOMATION IS REQUIRED TO MEET THIS CHALLENGE. RECENTLY, THERE HAS BEEN AN EXPLOSION OF INTEREST IN USING UNMANNED AERIAL VEHICLES (UAVS) AND OTHER KINDS OF AUTONOMOUS VEHICLES TO IMPROVE AGRICULTURAL PRACTICES. UAVS IN PARTICULAR HAVE BECOME UBIQUITOUS INSTRUMENTS ON FARMS ACROSS AMERICA, AND PRECISION AGRICULTURE IS EXPECTED TO BECOME THE LARGEST COMMERCIAL MARKET FOR UAVS IN THE YEARS TO COME. UAVS ARE USED AS EYES IN THESKY TO OPTIMIZE FARMING, AND THEY PROMISE TO PLAY A SEMINAL ROLE IN MAINTAINING THE RATE OF PROGRESS IN AGRICULTURAL EFFICIENCY. THEY ARE USED TO SUPPORT BETTER MANAGEMENT OF FARM INPUTS BY DETECTING, LOCATING, AND IDENTIFYING: NUTRIENT DEFICIENCIES, WEED DENSITIES, DISEASE, PEST INFESTATIONS, AND CROP WATER STATUS, ENABLING EARLY INTERVENTION TO PROTECT CROP YIELDS.ONE CRUCIALLY IMPORTANT APPLICATION FOR UAVS IN AGRICULTURE IS THE REMOTE DETECTION OF CROP STRESS. MANY FORMS OF STRESS MANIFEST IN CHLOROSIS, WHICH IS A YELLOWING OF THE FOLIAGE DUE TO A REDUCTION IN CHLOROPHYLL PRODUCTION, AND CAN BE READILY DETECTED FROM THE AIR USING A VNIR (VISIBLE + NEAR INFRARED) CAMERA. AERIAL IMAGING WITH VNIR CAMERAS MAKES IT POSSIBLE TO DETECT STRESS WITH EXCELLENT RESOLUTION AND COVERAGE RATES (THE AREA YOU CAN EVALUATE PER UNIT TIME). THIS IS WIDELY USED TO DETECT CROP NUTRIENT DEFICIENCIES AND OTHER FACTORS THAT MAY REQUIRE INTERVENTION. SUCH AERIAL IMAGING IS OFTEN HAMPERED BY CLOUDS, WHICHCAST SHADOWS ON THE GROUND THAT CAN INTERFERE WITH THE INTERPRETATION OF THE IMAGERY. AT ANY GIVEN TIME, APPROXIMATELY TWO THIRDS OF THE SURFACE OF THE EARTH IS COVERED BY CLOUDS, WHICH MAKES THIS PROBLEM QUITE SIGNIFICANT. AT PRESENT, PEOPLE MUST TRY TO COLLECT IMAGERY ONLY ON DAYS WITH CLEAR OR UNIFORMLY OVERCAST SKIES, OR RISK MISINTERPRETING THE IMAGERY AND, FOR INSTANCE, UNDER OR OVER-APPLYING FERTILIZERS AS A RESULT.IN THIS PROJECT WE AIM TO SOLVE THIS IMPORTANT OPEN PROBLEM BY DEVELOPING A SYSTEM OF MULTIPLE COOPERATING UAVS THAT WILL TRACK CLOUD SHADOWS AND COLLABORATE TO COLLECT SHADOW-FREE IMAGERY OF AN ENTIRE FIELD. THIS WILL SIGNIFICANTLY EXTEND THE RANGE OF CONDITIONS UNDER WHICH UAV-BASED IMAGING CAN BE CONDUCTED, AND MAKE IT FEASIBLE TO INCORPORATE UAVS AS A RELIABLE AND MORE INTEGRAL PART OF FARM AND NUTRIENT MANAGEMENT METHODOLOGIES. THIS PROJECT WILL EXPLOIT AND ADVANCE SEVERAL FACETS OF SCIENCE AND TECHNOLOGY. IN PARTICULAR, THIS PROJECT REQUIRES THE DEVELOPMENT OF NOVEL MULTI-OBJECTIVE, MULTI-VEHICLE CONTROL SCHEMES TO ENABLE A TEAM OF ROBOTS TO COLLABORATE AND SAFELY ACCOMPLISH A COMMON SET OF SOMETIMES COMPETING OBJECTIVES. TO MINIMIZE THE COMPLEXITY OF IMPLEMENTATION, THESE CONTROL LAWS NEED TO BE DESIGNED TO ENABLE THE SELF-CONTROL OF EACH VEHICLE, AND THEY SHOULD REQUIRE AS LITTLE VEHICLE-TO-VEHICLE AND VEHICLE-TO-GROUND INFORMATION EXCHANGE AS POSSIBLE. THE CONTROL LAWS NEED TO BE RELIABLE AND ROBUST. THAT IS, THEY MUST BE ABLE TO TOLERATE INFORMATION INACCURACIES AND COMMUNICATION DELAYS, ALONG WITH ENVIRONMENTAL PERTURBATIONS SUCH AS WIND GUSTS. THIS PROJECT ALSO INCLUDES THE DEVELOPMENT OF A SYSTEM FOR TRACKING CLOUD SHADOWS IN REAL-TIME AND PREDICTING THEIR MOVEMENT AND EVOLUTION USING ADVANCED NEURAL NETWORKS. DEVELOPMENTS AND PUBLICATIONS RESULTING FROM THIS PROJECT WILL CONTRIBUTE SIGNIFICANTLY TO THESE ACTIVE AND IMPORTANT AREAS OF RESEARCH.P { MARGIN-BOTTOM: 0.1IN; LINE-HEIGHT: 115%; BACKGROUND: TRANSPARENT NONE REPEAT SCROLL 0% 0%; }$174,182
· FY2020 · National Institute of Food and Agriculture