Delhi Technological University

Team UAS

Delhi Technological University

2010 AUVSI STUDENT UAS COMPETITION
Team UAS DTU Journal Paper

Abstract
This paper provides a description of the Delhi Technological Universitys (DTU) Unmanned Air System (UAS)
designed for Association for Unmanned Vehicle Systems International (AUVSI) Student UAS Competition 2010. The
teams UAS Bulbul is a modified Sig Rascal 110 R/C aircraft controlled by an indigenously developed autopilot
system. Bulbul uses a gimbal stabilized network camera to transmit real time imagery to its Imagery Station over a
Wi-Fi link for real time target detection. Safety is of paramount importance as Bulbul can be controlled by its
Ground Control Station over a 2.4 GHz wireless link as a Remotely Piloted Vehicle (RPV) and also by a 72MHz
Remote under full manual control.
This paper gives a detailed description of Bulbuls system along with the design rationale including the teams
objectives in building the system. The paper also describes Bulbuls flight operations and concludes with a
description of the testing that has been performed to show that the mission can be completed safely and
successfully.

Delhi Technological University

Team UAS

Table of Contents
Abstract

Table of Contents

1.0 Introduction

1.1

Team

1.2

Mission Requirements Analysis

1.3

Systems Engineering Approach

1.4

Systems Overview

4
5

2.0 Design and Component Overview

2.1 Aircraft

2.1.1

Airframe

2.1.2

Propulsion

2.2 Autopilot

2.2.1

Onboard Computer (OBC)

2.2.2

Sensors

2.2.3

Component Box

2.2.4

Command and Control Station

2.3 Imagery

2.4 Communications

12

2.4.1

Manual Flight Remote

12

2.4.2

Telemetry Systems

12

2.4.3

Imagery Downlink system

13

2.5 Power Systems

14

2.6 System Layout

15

3.0 Failure Mode Analysis

17

4.0 Operations

18

5.0 Conclusion

19

6.0 Acknowledgements

20

7.0 References

20

Delhi Technological University

1.

Team UAS

Introduction

1.1 Team
Delhi Technological Universitys Student UAS Competition team is comprised of an interdisciplinary group of
students. The team thoroughly analyzed the competition requirements and came up with a systems engineering
approach to design the UAS to achieve the competition objectives.
A modular approach to design was taken; all systems were designed in modules and were thoroughly tested during
development as well as integration stages. As a result the final system provided satisfactory results.

1.2 Mission Requirements Analysis

The AUVSIs Student UAS Competition simulates a real world scenario. The basic objective is to support a company
of US Marines conducting a patrol with Intelligence, surveillance and Reconnaissance (ISR). To meet this objective,
the UAS must be capable of following a set of pre-defined waypoints while identifying targets on ground. In
addition to this, the UAS must be capable of changing its flight plan while in flight. Lastly it should be able to search
for targets in a pre-defined area.
Throughout the mission, the UAS should operate without compromising the safety of its operators as well as that
of the environment in which it is operating.
The mission requirements are further analyzed in the various sub sections below where they are co-related to the
design decisions taken by the team.

1.3 Systems Engineering Approach

The project was divided into 3 phases: Conceptual Design, Preliminary Design and System Integration & Testing.
In the Conceptual Design phase, the team came up with threshold and objective requirements of Bulbul. The
previous years system was analyzed thoroughly based on its performance at the competition and various
shortcomings were identified. Also, the systems of the other teams were studied carefully before coming up with a
system for this years competition. Many potential systems were considered, with an overall score being assigned
based on the requirements. Finally, the system with the best overall score was chosen, with safety and reliability
being key considerations. A schedule was then drawn up and work was divided among the team members. In
Preliminary Design, individual modules were developed and extensive unit testing was carried out in the laboratory
as well as in field testing. This proved beneficial as system integration was made much easier, due to reliability of
the subsystems.
In System Integration & Testing, the successfully tested modules were combined together and performance was
verified in flight tests. Weekly team meetings were held to ensure that work was progressing on schedule; and for

Delhi Technological University

Team UAS

team feedback to be delivered. Regular design reviews were also helpful in detecting and resolving problems at an
early stage. They were also used as a good opportunity for carrying out quality control and safety checks.
1.4 System Overview
The teams primary focus for this year was to ensure the reliability of the avionics system, as well as to improve the
imagery and communications systems. The overall safety of the system was also of prime importance.
Bulbul is made up of 3 subsystems: Airframe, Autopilot and Imagery.
Bulbuls airframe is a modified Sig Rascal 110, a popular R/C aircraft manufactured by Sig Mfg. The modifications in
the aircraft made it possible to house both the autopilot and the imagery payload. The autopilot is responsible for
controlling the aircraft in autonomous mode. It communicates with the Ground Control Station over a 2.4 GHz WiFi link. The imagery system consists of a gimbal stabilized Axis 207 MW network camera, transmitting real time
imagery to the Imagery Station over a 2.4 GHz Wi-Fi link. The imagery and autopilot systems are completely
separate and use separate hardware for communication ensuring reliability. This approach also led to shorter
development time and ease of testing prior to integration in the UAS.

Bulbuls System Overview

Delhi Technological University

Team UAS

2. Design and Component Overview

During system design the following approach was adopted. For each component, such as the airframe, autopilot,
camera, batteries etc., key performance parameters were identified and used to compare components or
technologies used and a score was assigned. The highest ranking individual components were then assembled
together into an overall system and an overall score was assigned. Several such combinations were explored, with
performance metrics being based on competition requirements such as autonomy, image quality, flight time,
safety and development time.
2.1 Aircraft
The airframe being used by the team is the Sig Rascal 110 ARF. The airframe is powered by O.S. 160 FX, a 2-stroke
glow fuel based engine running a 17x6 propeller. The engine provides more than sufficient thrust and hence
allowed the team to test a broad variety of onboard systems (payload).
An airframe back-up with identical specifications as of the main airframe has also been prepared.
SPECIFICATIONS CHART FOR AIRCRAFT:
Wingspan:

110

Empty weight

11.2 lbs

GTOW

18 lbs

Engine power

3.6 hp

Endurance

40 min

Cruise speed

39 mph

2.1.1 Airframe
The team had used the same aircraft last year. Limited size and access to the payload bay were the biggest
problems faced last year. The team considered other available RC aircrafts and also a new self-made design.
However, the team decided against going in for a self designed custom aircraft because of lack of time to
prove its reliability levels. After a weighted evaluation, the team decided to stick to the current aircraft. Extra
accesses to the payload bay have been added and the component placement has been extensively redesigned
so as to comfortably fit into the available payload space.
The airframe has been tested over all the flight tests. Modifications and reinforcements to the fuselage have
been done after reviewing each flight test. The landing gear has also been strengthened so as to cope up with
the extra weight. A tensioned chord has been added which limits the maximum bending of the landing gear,
thus protecting the gimbal.
Foam has been added in the area around the firewall. This is done so as to ensure that if at all there is any fuel
leakage, the payload bay remains totally dry.

Delhi Technological University

Team UAS

The tips of the propeller blades have been painted to increase their visibility.
Another important safety feature that was added was the addition of programmable servos for the control
surfaces. The servos can be programmed for a fail-safe orientation in case of loss of RC signal.

Payload Bay

2.1.2 Propulsion: The team had to make an important decision whether to switch to an electric propulsion
system after the engine reliability issues in last years competition. Various possible systems were studied,
however, after the analysis the team came to the conclusion that to achieve the required endurance, an
electric system would not prove beneficial at all because of the weight of the batteries. Hence, the team
decided to stick with the glow-fuel powered 2 stroke engine.
The problems relating to engine starting were drastically improved after changing the orientation of the
engine. The previous orientation had resulted in repeated flooding of the engines cylinder.
The fuel tank has been upgraded and the aircraft can now give a 40 minute flight time.

17x6 Propeller

OS 160FX 2-stroke Engine

Differently sized propellers were tested during the flight tests.

Based on the performance, flight stability and the optimum speed for imagery, the team finalized on the 17x6
propeller.

Delhi Technological University

Team UAS

2.2 Autopilot
Bulbul employs an in-house autopilot which has been under development for the last two years and is one of
the major long term aims of the project. As the autopilot is a critical component for a successful mission, a
difficult decision which needed to be made at the beginning of the year was to choose between purchasing a
COTS system or going in for our own in-house system. It was observed at the competition last year that
despite having a COTS autopilot, many teams were unable to fly autonomously. Also, being under export
restrictions, it was difficult to obtain a reliable commercial autopilot in time for the competition. Hence, the
team decided to go in for development of the autopilot.
The autopilot can autonomously navigate through a set of predefined waypoints, while maintaining stable
flight. It communicates with the Command and Control Station (CCS) over a 2.4 Ghz Wireless link, can upload
waypoints dynamically, and fly in RPV mode.
The various parts of the autopilot are:
2.2.1 Onboard Computer (OBC)
The TS-7800 Linux based ARM Single Board Computer functions as the primary OBC in Bulbul. It is responsible
for Data Acquisition from Sensors, running the flight controls and navigation algorithms, and communicating
with the Ground Control Station to send telemetry and receive commands.

TS 7800

PIC 18F Microcontroller

Two PIC18F microcontrollers are used as secondary computers to offload computation from the TS 7800. The
first microcontroller is responsible for reading airspeed, avionics battery level and the instantaneous surface
deflections of all control surfaces. This last feature aided in control system development and can be used to
reconstruct the flight in a simulation later. The second microcontroller acts as a Servo Control Board (SCB) and
relays the required servo commands as generated by the flight controls to the servos.
Linux is a free and open source platform and gives soft real time capabilities. The autopilot was programmed
using C++, which is a proven language for such applications. The TS 7800 has a 500 MHz processor with 256
MB RAM, which is sufficient for the autopilot to perform in real time. The team had used the TS 7800 the
previous year as well and was satisfied with its performance.
Last year, a commercial SCB was used to control the flight surfaces. However, it was found to be unreliable
and had a slow response time. Hence, the team decided to develop its own SCB using a microcontroller, which
offered an additional advantage that a failsafe could be programmed into the SCB in case of failure of the

Delhi Technological University

Team UAS

primary computer. The SCB has been successfully tested to control up to 9 servos at 50 Hz. The avionics
battery level monitor was added as an important safety feature this year.
Extensive testing of the onboard computers was carried out and the results have shown that the system was
reliable and has worked well in flight.
2.2.2 Sensors
The Mti-G Attitude Heading and Reference Positioning System (AHRPS) from Xsens acts as the primary sensor
suite for the aircraft. It provides the roll, pitch, yaw, gyroscopic rates, accelerations, altitude as well as the GPS
co-ordinates along with groundspeed. The Eagletree V3 sensor is used to provide the airspeed.

MTI-G

EagleTree Airspeed sensor

The Mti-G was a replacement for the M-Nav IMU from XBow which was used last year. An accurate AHRPS is
the most basic requirement for successful autonomous flight. The M-Navs output had to be processed
through a Kalman filter algorithm which was not very robust, hence a switch was needed. Also, the Mti-G is a
complete solution which also avoids introducing an extra GPS module, in addition to the IMU. The AHRPS has
been observed in real time during flights and the telemetry has proven to be accurate.
2.2.3 Component Box
It was observed that one of the major flaws in last years system was the haphazard component placement
and wiring. Due to space restrictions in the payload bay, it was very difficult to troubleshoot the avionics and
the entire system had to be removed to effect repairs. This year, a component box was designed to hold the
flight computers, which could be easily disconnected and brought out without disturbing the rest of the
avionics, in case of an emergency. Also, if needed, reprogramming could be directly carried out by opening the
lid of the component box without taking it out, making the process very easy. Another major advantage was
that the component box provided Electro Magnetic Interference (EMI) shielding for the sensitive components
inside it, which was a major cause of concern. The component box also made the component placement
process easier, being a modular subsystem.
2.2.4 Command and Control Station
The Command and Control Station is the interface on Ground to the autopilot. It displays the flight telemetry,
avionics battery level, and instantaneous servo angles. It can be used to update waypoints, gimbal offsets, and
control system gains dynamically. It also contains a Virtual Cockpit feature so that the operator can have a
more intuitive idea about the current aircraft state. Lastly, a plugin has been developed which allows the CCS
to interface with Google Earth. Using this feature, the current location of the UAS can be tracked and plotted

Delhi Technological University

Team UAS

on a satellite imagery map of the terrain below the UAS. Warning alarms are generated if the UAS enters a NoFly-Zone, any flight parameter is out of a safe range, or if the battery level is too low. The CCS operator then
immediately informs the required personnel so that corrective action can be taken.
Previously, Java had been used to program the CCS Interface. However, this resulted in excessive usage of
RAM (up to 250 MB) and the processor (100% utilization), while having a slow update rate. This year, Python
was chosen as the programming language, and the improvements in performance were huge. The program
now uses only 14 MB RAM,has less than 40% processor utilization and is updated at 5 Hz. The communication
protocol was also changed from serial to TCP/IP. This resulted in a more reliable link and also increased the
data throughput rate from 9.2 Kbps to 54 Mbps.
The CCS has worked well in the flight tests conducted so far and is expected to perform reliably at the
competition.

Bulbuls CCS Interface

2.3 Imagery
Bulbuls imagery system consists of the Axis 207MW Network Camera, a two axis gimbal, and wireless routers to
provide a communication link to the Imagery Station. The Imagery Station is composed of two computers on a
network, which carry out the required processing on ground and provide the targets location and characteristics
to the US Marines in real time so that immediate action can be taken.
One of the key objectives this year was to completely overhaul the imagery system, so as to make it capable of
transmitting video in real time and providing images of sufficient quality to permit autonomous target recognition.
The most important requirements for the camera were: good image quality, a high resolution, and small weight
and size to enable the use of a gimbal to provide stabilization in flight. The Axis 207MW was deemed to be the
most suitable choice which fulfilled the above criteria.

Delhi Technological University

Team UAS

AXIS 207 MW
The previous years system had been based on an analog camera using a 2.4 GHz analog transmitter. It was found
that the downlink was prone to interference and the image quality was very poor. Hence, the camera as well as
wireless technology needed to be changed. Also, after analysis of flight videos with the above camera as well as a
Nokia N73 mobile phone, the team was convinced that gimbal stabilization was essential. A number of high
resolution digital cameras were considered, however the size and weight were prohibitive to the use of a gimbal.
The Axis 207 MW was the perfect compromise, providing 1.3 Megapixel resolutions while at the same time
weighing less than 190 grams. A network camera also allowed the imagery system to be isolated from the
autopilot, as the camera is on a separate wireless network from the telemetry, and will continue to work in case of
autopilot failure.
The video is received by the Image
Acquisition Node at the Imagery
Station. It is displayed in a window
inside a Graphical User Interface
(GUI)
built
using
National
Instruments Labview [1]. The
Imagery Spotter Team observes the
video during the mission, and if a
potential target is found, clicks on a
button to save the image. The saved
images are transferred automatically
over a Local Area Network to the
Image Processing Node, where
immediate processing is carried out
in a Labview program.
GUI for Imagery Spotter Team
The video is received by the Image
Acquisition Node at the Imagery
Station. It is displayed in a window inside a Graphical User Interface (GUI) built using National Instruments
Labview. The Imagery Spotter Team observes the video during the mission, and if a potential target is found, clicks
on a button to save the image. The saved images are transferred automatically over a Local Area Network to the
Image Processing Node, where immediate processing is carried out in a Labview program. A three step approach
was used in image processing . The First step consists of identifying the target location in the snapshot sent from
the Image Acquisition Node. This step relied on the diffidence of HIS color value of targets and background green
background for extraction of targets. This extracted target image was then scaled to a particular size which was
then used in the second step of processing. Next, the scaled target was analyzed to find the color of alphanumeric
and background of targets. The final step was the most challenging one as there were various distortions in the

10

Delhi Technological University

Team UAS

image due to high speed and jitter. Filtering and image reconstruction methods were used extensively (after
converting the image to two bit format using HSI values found in first step) to get the best possible image for
neural analysis of targets shape and alphanumeric. Target characteristics identified in these steps were then saved
in a text document, as well as displayed on the screen. At the same time, the CCS sends the GPS co-ordinates of
the aircraft to the Image Processing Node, where it is tagged along with the image. The targets location is then
calculated and displayed along with the other characteristics.

GUI for Image Processing Node

Processed Image
Captured Image
The above system has been successfully tested and has performed well in multiple flight tests. The gimbal has
been tested with extreme flight maneuvers and can easily keep the camera stabilized. The team used many flights
to determine the optimum focus and resolution settings for the camera.
Also, Labview has proved to be extremely fast and efficient for image processing. All the above factors have lead to
increased confidence in the imagery system and the team expects to do well at the competition this year.

11

Delhi Technological University

Team UAS

2.4 Communications
One of the objectives this year was to ensure a reliable communication link with the ground station.
Bulbul employs three communications links:
GHz Wi-Fi link for telemetry and receiving commands from CCS using a wireless router
1.

2.4 GHz Wi-Fi link for imagery downlink using a separate wireless router on a different channel.

2.

A 72 MHz PPM Remote for hard cut over i.e. for switching over to manual flight.

Both imagery and telemetry communication systems use 2 antennas each in vertical and horizontal polarization to
ensure communications link in all possible orientations of the aircraft.
2.4.1 Manual Flight Remote:
A major issue last year was presence of glitches in flight servos. After extensive testing, the team narrowed down
the cause of glitches to two reasons:

Presence of Unshielded DC-DC converters for powering various avionics systems

72 MHz PPM remote

It was discovered that a PPM system was itself prone to interference whereas the DC-DC converters were the main
cause of interference. The combination of two led to glitches in the flight servos. Therefore the team shifted to a
PCM based 72 MHz remote. The team also experimented with a 2.4 GHz remote but due to extensive testing using
a spectrum analyzer, it was realized that the remote jumped on every possible frequency in the 2.4 GHz band as it
was based on a Frequency Hopping circuit. Further testing revealed that the remote occasionally jumped on the
telemetry and imagery bands. This could lead to packet loss during flight and would be a potential safety hazard.
Hence the 2.4 GHz remote system was not used.
2.4.2 Telemetry system
The telemetry system evolved from Maxstreams 24Xtreme serial modems. During testing, the team found that
the serial modems were unable to handle the large data throughput and that led to slow update rate at GCS and
loss of data packets. It was primarily due to the fact that modems had low output power (50mW). Thus the team
shifted to a TCP/IP based communication system which used wireless routers between the ground station and the
aircraft. TCP/IP is a very reliable protocol with many additional built in features that ensured the reliability of
communication system. The system is capable of automatic reconnection in case of connection loss.
On board the aircraft, the system uses a 5.5 dbi omni-directional antenna in horizontal polarization while a 2.2 dbi
omni directional antenna is used in vertical polarization. This was done when it was observed that aircraft mainly
lost contact during extreme roll manoeuvres.
A major advantage of using routers is that the range of the system can be increased by just adding more routers
configured on the same SSID in the CCS network.

12

Delhi Technological University

Team UAS

The router based system has been tested on ground and as well as in flights. The system shows good performance
in a non-urban environment and marked improvement over serial modems in an urban environment (university
campus).

2.4.3 Imagery Downlink system

As Bulbul employs a network camera for real time imagery, two choices were available with the team for imagery
downlink system:

Using Axis 207 MWs inbuilt wireless router with a suitable amplifier

Using a separate wireless router for the camera system

Among these, the team first experimented with the cameras in-built wireless along with a 750mW amplifier. At
the Imagery station, a wireless router with an 8dbi omni directional antenna was used to establish the wireless
network to which the camera connected wirelessly.
Many problems were experienced with this system, with the most significant being the lag in receiving imagery at
the Imagery Station. After extensive testing, the team concluded that the cameras wireless was unable to handle
the throughput at high resolutions. The low resolution imagery would render the UAS unusable at high altitudes.
Hence this system was dropped and team decided to use a separate wireless router similar to the one used for
telemetry for the imagery system as well. A similar router was employed at the Imagery station. The router system
offered many significant advantages including ease of setup and debugging. Like telemetry, the Imagery
communication system also used two antennas in perpendicular polarizations to ensure link with Imagery Station
in all possible flight orientations. On board the aircraft, the system uses a 5.5 dbi omni-directional antenna in
horizontal polarization while a 2.2 dbi omni directional antenna is used in vertical polarization.
As the team was using both imagery and telemetry communication systems on 2.4 GHz, mutual interference was a
major concern. However, testing revealed that both systems used completely different channels. The telemetry
and imagery communication system have been tested together in a variety of flight tests and have never shown
any interference.
It might have been a better choice to use a 900 MHz system for telemetry and use 2.4 GHz system for imagery as
this would have increased the range of telemetry and reduced chances of interference; however 900 MHz is not
available for use by general public in India as it is used in GSM cell phone networks.

13

Delhi Technological University

Team UAS

2.5 Power System

For powering the avionics, Bulbul uses one 14.8 V 3.3AH Lithium Polymer battery. The flight servos are powered by
another 11.1V 2.1AH Lithium Polymer battery. To provide a regulated power supply to various avionics subsystems, Battery Elimination Circuits (BECs) are used. Castle Creations programmable BECs can provide regulated
output voltage between (4.8V to 12.5V). All avionics systems run at either 5V or 12V and hence 2 BECs are used.

Switch Board
Last years power system was based more on theoretical knowledge than testing and experience. It was discovered
that un-shielded DC-DC Converters used for power distribution were a major source of EM noise in the aircraft and
were causing glitches in the servos. The team experimented with shielded DC-DC Converters and BECs in parallel.
The two systems were rigorously tested in lab before being used in any flights. Current flowing in all wires was
observed using CRO for signs of EM noise. Both the systems passed teams final quality checks, however BECs
were finally used for power distribution as they were 6 times lighter than shielded DC-DC converters.
To decide the battery rating the following POWER CONSUMPTION TABLE was drawn
S.No

Components

Quantity

Power Consumption (Watt)

1.

TS-7800 Single Board Computer

5.0

2.

Wireless routers

16.0

3.

Inertial Measurement Unit

0.92

4.

PIC microcontroller boards

6.0

5.

Camera

4.0

Total

31.92

14

Delhi Technological University

Team UAS

Image courtesy:

Battery eliminator circuit

ThunderPower LiPo Battery

Therefore, 32 watts of power is required to power the UAS for a one 1 hour flight. Based upon previous experience
of being conservative with factor of safety for power consumption and the teams desire to conduct multiple
flights before changing or charging the battery, the team decided to use a 14.8 Volt 3300mAh Thunder Power Li-Po
Battery.
A new feature added year is a central power control board to selectively switch on & off systems during testing
and debugging to save power. Also an avionics startup procedure has been established between the Ground
control team and Aircraft team to ensure proper startup and shut down of the entire the UAS system.
The current power system also provides a backup power to flight servos. In case the flight servo battery fails in
flight, the servo power system automatically shifts to avionics battery and keeps the vehicle in control.
The avionics and servo battery voltages are available at all times at the Ground Control Station.

2.6 System Layout

Core stress was given on the placement and integration of the various components being used inside BulBul. The
fuselage/payload bay can be broadly divided into 3 volumes: Batteries, Avionics and Radio Equipment
The following factors were kept in mind while finalizing the component layout:
1.

The C.G. of all the added components put together should be on the natural C.G. of the airframe.

2.

Components/ ports which needed to be repeatedly accessed were placed and oriented towards the
payload bay access.

3.

EMI effects.

4.

Ease of removal of components under emergency scenarios

5.

Heating tendencies of the components.

15

Delhi Technological University

Team UAS

One of the key factors during component finalizing was also its physical dimensions whose limitations were
decided by the airframe.
The component placement was first planned on various 2-D 1:1 scale charts of the payload bay and CAD [2]
models.
However, to understand the actual wiring and the limitation while emergency removal/replacement of the
components, it was necessary to actually place them into the airframe.
To prevent damage to the airframe and to provide maximum flexibility to the avionics team to arrange
components as per their needs, a 1:1 scaled and precisely dimensioned mock-up of the entire payload bay was
prepared. Thus, only after each and every detail of the entire system had been worked out in the mock-up and had
been approved by each of the 3 sub-teams was the system finally transferred to the main airframe.
The team came up with an avionics layout procedure which is a step by step process to integrate the avionics
system with the airframe.
Secure and neat wiring is a key parameter in ensuring the safety of UAS in flight. It also aids in quick debugging in
case of any problems. Keeping these factors in mind, the wiring routes were designed during the preliminary phase
of the design. All wires are labelled to ease connections. All the connectors used are unique so as to prevent any
accidental connections. All wires to the control surfaces servos are twisted and have ferrite beads as filters to
prevent any glitches. All co-axial cables are also shielded.
All Ethernet cables are locked so as to prevent any accidental disconnections.

16

Delhi Technological University

Team UAS

3. Failure Mode Analysis:

Failures are divided into 3 categories:

Code Yellow :Recoverable with intervention of CCS

Code Blue : Recoverable with intervention of RC Pilot

Code Red

: Critical Mission Failures

A code yellow failure becomes a code blue failure if the code yellow intervention fails to recover the state of the
aircraft in less than 3 minutes or as decided by the safety officer. Code blue failures require intervention of RC Pilot
and the mission would no longer be executed autonomously. If judges permit, RC pilot continues mission
execution aided by CCS (if possible) and spotter. Code Red failures require immediate landing if possible or a crash
landing in the nearest safe location.
FAILURE ANALYSIS TABLE
Failure
Loss of airspeed

Large Navigation
errors
Loss of GPS in
flight

Most probable cause

High angle of attack
Sensor Failure
Flight servo failure
Crosswinds
Sensor Failure
Interference
GPS receiver failure

Loss of Altitude

Altitude hold loop not

working
Barometer failure
Flight servo failure

Loss of Telemetry
signal

Interference from other

sources

Loss of Imagery
signal

Interference from other

sources

Engine failure

Fuel flow to engine cut

off momentarily

Aircraft breaks in

Undue aerodynamic

Code Yellow
intervention
CCS commands the
aircraft to pitch down
and gain speed
The CCS corrects the
required PID gains
CCS commands Bulbul
to fly at constant
altitude and orbit
around current position
at 20 degree bank.
CCS resets the autopilot
which immediately
brings aircraft to level
flight and then
commands to gain
altitude overriding
navigation response.
Bulbul flies back to first
waypoint and orbits
around it till connection
is re-established.
Bulbul flies back to first
waypoint and orbits
around it till connection
is re-established.
N/A

N/A

17

Code Blue
intervention
RC Pilot takes over
mission execution

Code Red
Intervention
Bulbul crash lands

RC Pilot takes over

mission execution
RC Pilot takes over
mission execution

N/A

RC Pilot takes over

mission execution.

Bulbul crash lands

RC Pilot takes over

mission execution.
Mission success is
severely compromised.
RC Pilot takes over
mission execution.
Mission success is
severely compromised.
RC Pilot takes over and
lands the aircraft
Mission Failed
RC Pilot takes over and

N/A

N/A

N/A

N/A

Bulbul crashes

Delhi Technological University

flight

stress

Low battery
voltage

Avionics system
switched on for more
than expected time
Interference in 72 MHz
band

Loss of manual
flight control (in
case of code Blue
intervention)

Team UAS
tries to land
Mission Failed
RC Pilot takes over
mission execution

N/A

N/A

N/A

N/A

Bulbul executes
crash landing
failsafe mode and
crashes.

4. Operations
The flight operations were drafted after a detailed and meticulous analysis of the mission requirements and the
final outcomes expected. The key considerations were:

Safety: Of the flight crew as well as the airframe and its systems.

Portability: The entire system had to be made capable of being quickly and efficiently set up out of a
single container.

Set Up time: The setup of the entire system (Ground Control Station and the aircraft) had to be done
under the allotted time of 40 minutes.
The entire UAS mission has been divided into the following stages and further planning has been done based on
that.

M1: Collection of required equipment before leaving for the mission.

M2: Setup of the CCS and the airframe.

M3: Testing of the all the systems and communications.

M4: Operations during flight

M5: Post flight operations

Furthermore, responsibilities were allocated to the flight crew.

Flight Director: Responsible for the overall management of the mission. He gives the final go ahead for the mission
after ascertaining that the Ground Control Station team and the airframe team are ready for the mission. The final
decision under any circumstances would be made by the flight director.
Safety Officer: The safety officer oversees the entire mission to keep a check on potentially unsafe situations. To
avoid such scenarios, he is responsible to continually advise spectators and the flight crew.

18

Delhi Technological University

Team UAS

CCS and airframe heads: They are responsible for the setup of their respective systems. After the setup, they
oversee a quick system check-up and give a go-ahead to the flight director for the mission.
Spotter: When the aircraft is in the manual mode, he assists the pilot by providing him directions etc. He
continuously monitors the GPS location of the aircraft and warns the pilot if he is about to fly out of bounds.
Imagery Lead: He ensures an optimal setting for the imagery. He also loads the images which are to be processed
autonomously by the imagery program. At the end, he generates the Excel sheet which contains all the target
parameters.
Imagery Spotters: Their task is to observe the live imagery on the separate terminals and notify the imagery lead
whenever a potential target is spotted.
Tool-Man: He assists in rapid setup of the airframe by providing all the tools required for the assembly.
Team Liaison: The team liaison handles all the communications with the judges and the spectators during the
entire course of the mission.
The Flight Crew responsibilities have been so allocated that the mission can be performed with a minimum of 5
members (and pilot) and optimally with 10 members.
Each member of the flight crew has a detailed check-list for each phase of the mission. This ensures a quick and
sure implementation of all the procedures, while at the same time keeping the communication between members
brief and precise.

5. CONCLUSION
The entire procedure that has been devised has been improved and updated after all the flight tests carried out so
far. Further analysis also revealed the key components required for the mission and a back-up for the same has
since always been in place.
The mission operations also helped us to better understand the failures that could occur during its entire course.
The team has handled emergencies like fuel-leakage in the payload bay to a complete power shutdown on the CCS.
This helped us to come up with a fall-back plan in almost all kinds of imaginable scenarios.
The switching of systems from the main to the back-up airframe has been achieved in under 20 minutes.
At the time of writing the journal, the team had performed a total of 9 flight tests with over 25 flights in all. The
flight data logs reveal that the aircraft has been in air for over 3 hours.
During the course of these flight tests, we have been able to bring the setup time for the entire system to under 33
minutes (including the time to set up the tents).
The turnaround time (time between 2 consecutive flights) has been under 12 minutes. Mission testing and
simulations reveal that BulBul can complete the entire mission within 30 minutes

19

Delhi Technological University

Team UAS

6. ACKNOWLEDGEMENTS
We would like to thank the following members of Delhi Technological University for all of the invaluable
guidance and support that they have given our team:
Prof. P.B Sharma (Vice Chancellor, DTU)
Prof. R.K. Sinha (Dean, Industrial Research and Development)
Dr. D.S. Nagesh (Faculty Advisor)
Mr. N.S. Raghava (Faculty Advisor)

7. References
1.

www.ni.com/labview - National Instruments Labview

2.

www.solidworks.com - SolidWorks 3D CAD Design Software

20