Design and Development of an Intelligent UAV System with Obstacle Avoidance and Object Detection using Machine Learning

The usage of UAVs is gaining importance with the growing demand of fast and reliable delivery of products. There are different ways in which these vehicles can be built for different purposes. They are extensively used in different domains such as Delivery, Military, Mobile Computing, Agriculture and many more. This article explains one such way of building a prototype of self-driven aerial vehicle. This research provides insight on how the vehicle moves from one place to another based on coordinates such as latitude and longitude values. With the integration of Machine Learning and Computer Vision the vehicle can identify the type of object present in the scope. The distance sensors such as Ultrasonic and Lidar helps the vehicle to measure the distance of the obstacle present in front of it. The vehicle's ability of taking self-decision based on location and environment can be coded with the Python's opensource Dronekit library. The software such as Mission Planner and QGroundControl helps us to understand the vehicle's present status through the wireless telemetry. This prototype can be extended for making a commercial aerial project that can be used for various purposes such as surveying, delivery and other research advancements in drone technology. The research done is aimed to achieve the Sustainable Development Goals (SDG)- SDG 4, SDG 8, SDG 9, SDG 11.