You got a new drone! A shiny quad-copter, complete with a remote. You put thebatteries in the remote controller and start flying. As you fly, you realize that thedrone can go only a little distance, unless you are following it. That’s a littleannoying! Your quad-copter has very limited range!

This drone kit was designed as quick DIY toolkit to get one’s AR ParrotDrone 2.0 to interface with, and be controlled by, their mobile devices.The communication happened over cellular data, thereby removing theneed for expensive WiFi beacons required to facilitate, and communicatewith, the drone

4g

A number of commercial drones, meant for non-businessusage, have very limited range of communication. They oftentalk to the remote controller over WiFi channels.

For this, drones need to have an in-built WiFi modulethat can connect with other WiFi enabled devices toreceive commands for control.

Hello!??

Can’t hear you..

Given the battery constraints, however, the drone’s WiFimodule lacks significant range. As a result, you have twooptions if you want your drone to go afar:

Walk with the drone

This simple solution becomes challenging since it’s difficult to keep up with thespeed of the drone, while managing the controls like pitch and yaw of the drone.As a result, your attention gets distributed and you focus more on controllingand keeping up with the drone, instead of having fun with it.

Setup WiFi beacons

You plan the drone path ahead of time, and setup WiFi beacons so that they canact as communication relay for your drone. This method is extremely expensive,not to mention tedious and highly constrained.

Imagine setting up beacons everytimeyou want to fly the drone! It kinda takesthe fun out of it. And unless you’re adeptat dealing with wifi beacons and othersimilar devices, it may be really difficultfor you

CONSTRAINED

Every time you want to fly the drone tolarger distances, you have to know thearea at least, if not the path. Further, notall terrains will be suitable to setup theWiFi beacons

TEDIOUS

Even if you don’t mind the constraints and tediousness, theprice of the beacons will not be exactly cheap, not tomention the additional setup cost. And sensors and smallcommunication devices often have high failure rate in thewild (uncontrolled) environments

EXPENSIVE

In this problem, there are two areas where we could intervene:

Enhancing the WiFi reception of the drone.This is difficult given the proprietary natureof the device we were working with.

Replace the need of WiFi beacon withsomething more ubiquitous.

While people don’t carry wifi beacons around with them, there is one communication devicewhich almost everyone in an urbane scenario has - a mobile phone. We decided to figure a wayout for using mobile devices as remote controllers.

Our team created a small controller, that could be mounted on the drone itself. People could use their mobileinternet to talk to and give commands to this auxiliary controller, which then translated and passed them to thedrone. This is a highly simplified way of explaining the whole architecture we developed.

GPS

Microsoft Azure
IoT Hub

GSM

WiFi

WiFi

Mounted
controller

Mobile
phone

Server

Drone

We used AR Parrot Drone 2.0 for our design, development and testing. Due to the proprietary nature of the drone, itwas difficult to directly interface with the drone. So we designed an auxiliary controller that could be mounted on thedrone. Our mobile interface communicated with the auxiliary controller, which in turn communicated with the drone.

The auxiliary controller had the following components:

A GPS sensor

A WiFi module

A GSM module

Raspberry Pi 2(Running Windows 10 IoT):For providing a framework to
run all the other modules.

To get the location of the
drone for control

To communicate with
the drone

To provide the ‘mobile-data’communication channel, formobile interface to talk toauxiliary controller

We also developed a mobile application to demonstrate the end-to-endflow of controlling the drone through mobile interface as well. This was aUWP (Universal Windows Platform) application, that we customised formobile interface.

Mobile
phone

Server

GPS

WiFi

GSM

Our mobile application communicated the controls and directions to theauxiliary controller (mounted on the drone) through Microsoft Azure IoT Hub(?) over mobile data. For speed and performance, we used LTE (or 4G mobiledata) to communicate to the mounted controller, which interfaced with andcontrolled the drone.

Mounted
controller

Microsoft Azure
IoT Hub

This project was envisioned particularly around providing people the ability to live-stream events at theirconvenience. Given the numerous technical challenges and high-level nature of the problem, it had to be brokendown into simpler, more low-level problems, out of which, we chose the problem of restriction on communicationrange to work on.

We also faced some issues around the lag in communication with the drone. Forinstance, in the starting, it took the drone ~ 5 sec to receive a command andexecute it. For this, we had to redesign our architecture and re-structure thecomponents that we were using. We were able to bring the performance downto < 1 second.

> 5 sec

< 1 sec

We also aimed to provide more ‘intelligence’ to the drone, by building thecapabilities of obstacle detection and avoidance. We designed the system intheory, however, due to some constraints on time and resources, it couldn’t bematerialized completely.