Project Overview

Drone Pet seeks to utilize the increasing power of drones and video capture technology to provide a simple affordable solution to people who want to record themselves doing outdoor activities like skiing, hiking, etc. It provides a better view than having a goPro strapped to oneself, and can have its configurations adjusted for better angles of video capture. Our goal in developing drone pet was to make this technology cheaper, currently available off-the-shelf drones with similar capabilities go for upwards of $1500, and one of our biggest goals was to make drone pet as affordable as possible.

Project Milestones

As shown in our slides, we had planned four major milestones to cover over the course.

Milestone 1

Interfacing the drone with ROS and get it to fly semi autonomously in predetermined paths

Milestone 2

Set up the our AR parrot drone to be able to detect tags and provide meaningful data that we can use to set up the tracking algorithm. We did this using the inbuilt tag detection, and no openCV code was utilized.

Milestone 3

PID controller. To get fast/slow modes of adjustment so we can have slow (walking, hiking) mode and more aggressive mode for running etc.

Milestone 4

Autonomous Tag following. Getting everything together.

Milestone 5

Bug fixes.

Future Work

In the future, though outside the class, we plan to put a board on the drone itself so it that we can do all computations on the drone itself and not have to use a computer.

Also, we might also look into making Drone Pet into a ROS package where you can set up the autonomous navigation system for any drone that can interface with ROS and not just the AR Parrot.

Team Members

Aditya Bansal https://github.com/maddymanu

Dhruv Kaushal https://github.com/dhruvrrp