This project, sponsored by Boeing, aims to enhance the safety and reliability of unmanned aerial vehicles (UAVs) by successfully implementing a camera-based obstacle detection and avoidance system, utilizing computer vision, machine learning, and artificial intelligence to navigate challenging environments and contribute to the broader application of UAVs.
Sponsored by: Boeing
Team Members
Caleb Houser Brett Volker Curtis Young Mohammad Nur Amir Hakimi Bin Mohd Azmin Eduardo Casiano Parth Sureshbhai Seyed Sam Naemi Torshize Tyler Holman
Instructor: David Cubanski
Project Poster
Click on any image to enlarge.
Project Video
Project Summary
Overview
The rapid evolution of drone technology, particularly in smaller, more agile UAVs with extended operational ranges, has spurred the need for enhanced safety measures. This report encapsulates a project sponsored by Boeing aimed at augmenting the safety, predictability, and reliability of unmanned aerial vehicles (UAVs) through the implementation of an innovative obstacle detection and avoidance system.
Within this project, a focus is placed on integrating a camera-based object detection system using cutting-edge technologies like computer vision, machine learning, and artificial intelligence. The overarching goal is to adapt existing obstacle detection software and embed it into UAVs, facilitating autonomous navigation in challenging and dynamic environments.
Boeing’s sponsorship not only supports this endeavor by providing practical exposure for students but also aligns with their commitment to advancing autonomous solutions. This project’s successful implementation of a robust obstacle avoidance system stands to significantly enhance UAV capabilities, contributing to safer and more versatile applications in various industries.
Objectives
The primary objective of this project is to enhance the safety, predictability, and reliability of unmanned aerial vehicles (UAVs) through the integration of a camera-based object detection and avoidance system. Utilizing cutting-edge technologies such as computer vision, machine learning, and artificial intelligence, the goal is to adapt and implement robust obstacle detection software into UAVs, enabling them to navigate complex environments autonomously and contribute to advancing the practical applications of UAV technology. This project, sponsored by Boeing, aims to provide students with hands-on experience while aligning with Boeing’s efforts to implement autonomous solutions for safer and more reliable aerospace systems.
Approach
Our finalized solution integrates a camera-based object detection and avoidance system comprising components such as the NVIDIA Jetson Nano, Pixhawk 6C flight controller, Intel RealSense D435 camera, HolyBro X500 V2 Quadcopter, Endurance 5000 mAh Lipo Battery, the RadioMaster R81 Receiver, and RadioMaster TX16S radio transmitter.
The Intel RealSense camera captures stereoscopic data of the drone’s surroundings to be processed by the Jetson Nano. The Jetson Nano uses machine learning algorithms to assess collision risk and issue avoidance commands based on object type, size, and distance. The Pixhawk 6C flight controller, receiving commands through the MAVLink communication protocol from the Jetson Nano, adjusts motor speeds and direction to avoid collisions.
This collaborative integration of the camera, Jetson Nano, and Pixhawk 6C forms a robust system to enhance safety and reliability in UAVs.
Outcomes
The project reached a stage where it provided a solid foundation for the upcoming Capstone Design group, potentially sponsored by Boeing, to take over our portion of the project. Although not perfect, the project created a strong integration between the main parts. Put another way, the project’s next phase might be able to build on what we started and finish off any parts that were left undone because of schedule conflicts or improperly completed tasks from earlier in the project.