This document describes a thermal mapping drone designed to detect temperature differences on rooftops. The drone uses an infrared sensor and camera to map temperatures and identify poorly insulated areas. It transmits sensor and image data to a ground station that displays the information to users. The team successfully flew the drone and detected temperature variations in testing, though flight time was shorter than planned. The document provides details on the drone's components, data transmission, user interface, and testing results.

1. Thermal Mapping Drone
Jamyang Tenzin, Dylan Fallon, Stefan Totino, Jason Fellows
Faculty Advisor: Prof. Joseph Bardin
Department of Electrical and Computer Engineering
ECE 415/ECE 416 – SENIOR DESIGN PROJECT 2015
College of Engineering - University of Massachusetts Amherst
SDP15
System Process and Overview
Results
Acknowledgements
Abstract
Our system is a remote controlled drone capable
of detecting significant temperature difference,
with respect to the average temperature of an
area below its flight path. Our intention is for
homeowners to use the device as tool to scan
their rooftop and identify areas of poor insulation.
Utilizing emerging drone technology is a safer
alternative to the current hand held instruments
used for this type of application. Data from an IR-
sensor complemented with a JPEG camera
creates an intuitive map users can analyze.
System Block Diagram
• QuadCopter: APM 2.6 with barometer,
gyroscope, GPS & accelerometer sensor for
flight Control with Lipo Batteries, electronic
speed controller (ESC) and motors
• IR & Visible Light Cameras: MLX90620 IR
Sensor, Adafruit TTL Serial JPEG camera
• UltraSonic Sensor for height measurement
• Radio Link: Nordic NRF24L0 Transceiver
• 3D printed Cover from M5
• User Interface: Java/JavaScript/HTML Web
Application
Specifications
Specification Goal Actual
Weight < 2000g 1360g
Flight time 10 min 5 min
Power Consumption < 250 W 239.5 W
Cost < $500 $1,100
Max flight altitude 20 ft 50 ft
IR-sensor Temperature
range -50C to 300C -50C to 300C
IR field of view (FOV) 60° 60°
Radio link range >100 ft 300 ft
• Pilot navigates to scan the roof
• If temperature difference of 5˚C with
respect to average is found, LED light
Signals pilot to stop
• Infrared and distance readings are
transmitted from drone to receiver
• JPEG picture is taken and transmitted
• Data is received and displayed
• LED light shuts off signaling pilot to
continue scanning
We would like to thank Prof. Joseph Bardin for
advising us on this project. Special thanks to Prof.
Christopher Hollot, Prof. Christopher Salthouse &
Fran Caron for your guidance.
• Successful flight with full load
• Temperature difference detection algorithm
• Transmission of IR data and 12KB JPEG
over radio link
• USB driver on receiving end feeds incoming
data to Java application
• Data processed and displayed via a web
browser
Jamyang Tenzin, EE
Stefan Totino, CSE
Prof. Joseph Bardin
Jason Fellows, EE Dylan Fallon, EE

2. Cost
Drone Radio Link
IR Sensor
Receiver Application
Experiment
Arduino Mega
Power Module
Flight Controller Motor Unit
IR- Sensor Unit
Rx Tx
Radio Controller
Power Distribution
signal
signal
• Flight Controller APM 2.6 takes input signals
from Radio Controller and Outputs required
signals to Electronic Speed Controller (ESC) to
do proper flight manuever
• Electronic Speed Controller (ESC) converts DC
current to AC current depending on the signal
coming from APM 2.6
• 1000KV brushless motor takes AC current from
ESC to power itself with maximum total thrust
2480g
• 1800mAh Lipo Battery is used for powering both
motor and flight controller
Qty Part Development Production
1Drone 330mm Frame $11.25 $5.00
41000KV motors $66.60 $60.00
4motor accessory $8.36 $6.00
1Radio Controller $69.97 $69.97
4Carbon Fibre Propellors $9.12 $9.12
11800 mAh Lipo Battery $15.90 $15.90
1APM 2.6 Flight Controller $159.99 $159.99
4Electronic Speed Controller $103.96 $103.96
1IR Sensor $87.60 $50.00
1Arduino Uno $13.00 $13.00
1Arduino Mega $29.00 $29.00
1Transciever $12.00 $4.00
1camera $ 39..99 $39.99
1ultrasonic sensor $10.99 $10.99
1radio telemetry $23.86 $15.00
4Drone legs $8.00 $2.00
1Lipo Charger $21.35 $21.35
1Lipo Charger adaptor $13.00 $10.00
1shock absorber $10.00 $5.00
Total $713.94 $630.27
• 2 Nordic NRF2401+ transceivers
• Transmits data 32 byte packets at a time, IR data
is permitted to transmit upon specified
temperature difference, camera data is
transmitted right after
• Data rate of 1 Mbps @ 2.4GHz
• Arduino Mega controls transceiver on drone,
Arduino uno controls transceiver at receiver
application
• Arduino Uno feeds incoming data into PC’s USB
port
• Processing IDE mini app reads from port and
writes data to a text file
• Text file contains 1 byte which represents drones
height above target, 64 bytes which represent
temperature readings, and about 12,000 bytes
which represent a JPEG image file
• Java program creates a .jpg file out of the 12,000
image data bytes and sends the other 65 bytes up
to the client side program using an HTTP API
• Client side program uses JavaScript/HTML/CSS to
create a colored grid out of the temps, overlay the
grid over the .jpg file and display the output in a
web browser
• For our experiment we setup a section of our room
to try to detect temperature differences. We laid a
grid of the field of view of our IR sensor on a wall.
We then put various objects of significant
temperature difference’s on the gird and looked to
detect them. The IR sensor was then moved
further away to check the accuracy from different
distances. In conclusion we were able to detect
temperature differences with good accuracy,
accuracy decreased as we increased the field of
view for the IR sensor
• MLX 90620 contains 64 infrared pixels.
• Utilizes a Proportional To Absolute Temperature
(PTAT) sensor to compare objects in its field of view
to the ambient temperature of the chip itself.
• Outputs 16x4 array of temperature readings
• 1Hz refresh rate
• IR and PTAT data stored in internal RAM,
accessed by Arduino through I2C interface