This project idea is to develop an intelligent bugs mapping and wiping (iBMW) robot to perform pest population spatial distribution and “surgical precision spraying” for pest wipeout. The iBMW is an affordable (less than $1,000) robot-driven robot, which has a Turtlebot 3 as the robot’s brain and an unmanned ground vehicle serving as the work platform. Based on the design, the robot will be able to recognize and classify the Navel orangeworm by using deep learning neural networks. In addition, several iBMWs can work in the field together in swarming mode day and night, so that it can realize temporal and spatial bug mapping. As a result, mapping can determine which areas are at the greatest risk and whether wiping treatment is needed by iBMWs.
University: UC Merced
ARI
Aerial Research Intelligence (ARI) is a service that allows search and rescue personnel to expand their options for locating missing persons in a more efficient manner using drone technology and machine learning capabilities. ARI can be used with any small unmanned aerial system that autonomously searches the area around an initial planning point and quickly processes aerial image data to detect people using a trained neural network. Using RGB and thermal video data from sensors on board the drone platform, ARI is able to determine images and GPS locations that indicate the presence of a missing person, which greatly speeds up the human intensive activity of reviewing footage.
Unmanned Ground Vehicle for Water Leak Detection
The Unmanned Ground Vehicle (UGV) project is the future of the next generation’s agricultural gadgets. Instead of the past’s archaic methods of manually searching for leaks with acoustic procedures, the UGV project hopes to replace the previous methods with a more strategic appliance – a powerful camera attached onto a fully autonomous vehicle. Utilizing a combination of dynamic image-processing techniques and mobilization, this rover-camera duo is able to autonomously navigate through way-points and detect pipe leaks more frequently, efficiently, and accurately than a field worker would. Powered and guided by the Mission Planner program and the Pixhawk Autopilot System, the rover is capable of decreasing overhead costs, and most importantly, aiding in the preservation of water. The project’s state of the art features are distinguished by its data collection platform, algorithm design, and user-friendly interface.
Low Cost Scientific Data Drones for Enhanced Melon Productivity and Security
The SmartMelonDrone project will use low-cost unmanned aerial vehicles (UAV) to help manage both the quantity and quality of melon produce. The platform, capable of multispectral imaging, together with image post-analysis software, will support field management, including nitrogen stress detection, water stress detection, pest monitoring, yield estimation, small animal activities, etc. Firstly, real-time imagery with high spatial resolution (centimeters) can be acquired for growing melon by RGB, near infrared (NIR) and thermal infrared (TIR) cameras. Based on this analysis tool, information about nitrogen stress, water stress, and status of crop, weed, insect and disease can be extracted for optimized fertilizer, irrigation, precision application of insecticide, fungicide and herbicide, respectively. Meanwhile, grazing animals are also monitored to prevent introducing pathogenic bacteria into the soil. In addition, pre-harvest and harvest yield estimations will be determined for production decisions.
Small, Low-Cost Unmanned Aerial Vehicles for CAL FIRE Reconnaissance
Wildfires are a major part of California’s ecology and take a large amount of resources to monitor, contain, and ultimately suppress. Cal Fire is the state entity that is responsible for suppressing wildfires in California. Operations help improve the ecology of the local habitats by protecting rare and/or unique ecological resources, as well as protecting human property. Air-fighting resources such as fixed and rotary winged aircrafts are often used in fire suppression efforts. However, these tools are expensive to utilize and sometimes pose safety concerns such as pilot fatigue and low visibility flight. The goal of this proposal is to reduce the use of full-scaled crew-carrying aircraft by using small Unmanned Aerial Vehicles (UAV) in fire monitoring operations. Their project will create an UAV that could provide 24-hour monitoring to reduce the cost and increase the safety of wildfire monitoring. This would allow for traditional aircrafts that monitor fires to be used for different missions (e.g. water drops, short hauls, or resupply).