Project
# | Title | Team Members | TA | Documents | Sponsor |
---|---|---|---|---|---|
59 | QR code tracking drone |
Alexander Amiel Angelos Guan Umer Belagam |
Chaitanya Sindagi | design_document1.pdf design_document2.pdf final_paper1.pdf final_paper2.pdf presentation1.pptx proposal1.pdf |
|
Team Members: Umer Belagam - belagam2 Alexander Amiel - aamiel2 Angelos Tingrui Guan - tguan2 Problem Statement: Many drones use a controller, like a game controller, to fly it. This is usually enabled with one’s phone to keep track of location and to be used as a camera to see where the drone is looking. What happens if the phone dies? if someone gets a call/notification, this can exit the camera mode from the phone/controller. What if the drone is locked onto a person, but it doesn’t know which one to focus on? Solution Overview: Use a unique pattern (Ex. a QR code or a simple unique shape) for the user to hold in order to control the drone. This would solve the problem where there are multiple people in sight and the drone may follow the wrong user. The drone would detect the pattern and follow it from a fixed distance and at a direct angle. When the pattern is not detected, the drone would halt in its current position and display warning with LED and alarm on board. The user can then hold the pattern in front of the camera for the drone to detect and follow. Potentially use rotatable camera for better tracking and more degree of freedom Use different patterns to indicate different commands like flying to the ground and shut down or potentially changing the distance the drone would keep from the pattern. Give priority orders to the commands to avoid confusion if multiple patterns are detected. Solution Components: - Sensor Subsystem - Flight/navigation control: GPS, accelerometer, gyroscope and magnetic sensors to keep the drone flying upright and stay in position when it needs to. - Obstacle avoidance: ultrasonic sensors on each side - Object detection: camera - Power Subsystem: Lithium ion batteries and a USB charging and discharging control circuit. - Electronic Speed Control and motors: 4 BLDC motors and 4 ESCs to control each of them - Camera Mount: 3d printed camera mount for the raspberry pi camera with servo controlled tilt Criterion for Success: The drone can detect the pose & distance of the QR code The drone can successfully track QR code’s trajectory within some error boundary Similar Projects and Competitors: The DJI Gesture tracking drones already presents a solution for this that tracks a drone using different hand gestures. They don’t always work reliably and the drone can get confused on which person's hands to track if there are multiple people. Questions for the TA: Are there any consumer drones/flight controllers you recommend for us to look at - Should we buy a ready made flight controller with all the sensors and a microcontroller to use those inputs for controlling the motors or should we make our own flight controller? - Any other method for tracking aside from a QR code like pattern. |