PROJECT TITLE:

Design and Implementation of

an Autopilot for a UAV

SUPERVISOR:MR. E. Zingapeta

STUDENT NAME SIN PROGRAM

Richard Aaron Mvula 21168956 Aeronautical

Lucas Banda 21171241 Aeronautical

Lusaka Simbeye 21159654 Aeronautical

1.0
Introduction/Synopsis/Background
UAVs-Unmanned Aerial Vehicles have been around longer than one might think. The
first UAV, the Hewitt-Sperry automatic aircraft, dates back to the first world war.
However, it is only more recently that the UAVs on the market have reached a state
of practical use. As with many other technical systems, The introduction of light
weight structures and dependable, cheap computers made it possible to make
simple UAVs work satisfactory.
 The rapid progress in energy storage, sensors, actuators, and computing has
significantly enhanced the development of Unmanned Aerial Vehicles (UAVs),
enabling them to perform complex tasks with greater efficiency [1]. UAVs are
particularly valuable for executing tasks in environments that are hazardous or
difficult for humans to access.
 Generally, UAVs have high levels of autonomy than typical aircrafts.
 1.1 Problem Statement
 Most of the existing autopilot systems lack the ability to detect and avoid turbulent
weather conditions and this usually leads to most of the UNMANNED AERIAL
VEHICLE’S (UAV’s) crashes.
1.2 Motivation/Justification

 The designing of an autopilot system with the aforementioned capability

will help reduce the number of aviation accidents due to UAVs by
approximately 3-5%.
 It will also prevent the owners of the UAV’s from making financial losses.
 As was noted by the Federal Aviation Authority(FAA) from the study
conducted in 2010 regarding UAV accidents, out of 297 UAV accidents 3%
were related to weather turbulence. However, this data comes from large
military UAVs such as the predator and gives no sense of the numbers of
the reported accidents for small unmanned aerial vehicles.
 Furthermore, the reduction in the number of accidents will also save
human lives in places where UAV accidents would have occurred.
1.3 Aim
 To design and implement an autopilot system that gives an UNMANNED AERIAL
VEHICLE(UAV) the capability to detect and avoid turbulent weather conditions.
1.4 Objectives
 To conduct a literature review on the history and development of autopilot
systems.
 To Design an autopilot system that enables the UAV to detect and avoid
turbulent weather conditions .
 To enable the UAV to navigate autonomously through a predetermined waypoint
list.
 To enable the UAV detect and avoid obstacles.
 Testing and validation of the autopilot system to meet performance
specifications.
1.5 Project Scope
 Designing an autopilot system.
 Implementing an autopilot system
 Fabrication of an autopilot system
 Software Development: Development of the autopilot software, including
the flight control algorithms, navigation system, and communication
protocols.
 System Integration: Integration of the autopilot hardware and software
with the UAV platform.
2.0 Literature Review
 Autopilot systems are central to enabling unmanned aerial vehicles (UAVs) to perform autonomous
operations, from navigation to mission-specific tasks. This review synthesizes advancements,
methodologies, and challenges in autopilot design, with a focus on academic and industrial
contributions, including key book sources.
 Control strategies for UAV autopilots balance stability, precision, and adaptability. These include the
following
 Fuzzy Logic Control: Fuzzy logic is widely adopted for handling nonlinear dynamics and
environmental uncertainties. Valencia-Redrovan et al. (2015) designed a fuzzy logic-based autopilot
that minimized computational processing while maintaining stability, altitude, and acceleration
control . However, studies comparing fuzzy systems with PID controllers note that modified PID
designs often achieve better robustness and settling times .
 PID Controllers: Traditional PID controllers remain popular due to their simplicity, but their transient
response during altitude maneuvers can be suboptimal. Hybrid approaches, such as phase-lead
compensators, improve performance by addressing these limitations .
 Adaptive Systems: Recent advances in adaptive control, as discussed in *Control of Autonomous
Aerial Vehicles* (L’Afflitto et al., 2024), emphasize underactuated systems for scenarios with partial
actuator failures or environmental disturbances. These systems leverage modular architectures for
flexibility in mission-critical applications .
2.0 Literature Review
 Modern autopilot systems require seamless integration of hardware and
software:
 Modular Hardware: ARM Cortex M4-based designs, such as those implemented
by von Eichel-Streiber et al. (2020), use distributed CPUs for safety and
scalability. Each hardware node operates independently, enabling fault tolerance
and expandability .
 Simulation Tools: Tools like Software-in-the-Loop (SITL) and Hardware-in-the-Loop
(HITL) simulations are critical for validating algorithms. NASA’s Reflection
Architecture exemplifies a component-based framework for real-time embedded
systems, allowing plug-and-play testing of autopilot modules .
 Open-Source Ecosystems: Platforms like ArduPilot and PX4 dominate UAV
development due to their compatibility with ROS, MATLAB/Simulink, and
MAVLink protocols. These systems support companion computers for advanced
tasks like real-time image processing .
2.0 Literature Review
 Reliable communication and fail-safe mechanisms are critical:
 Sensor Fusion: LiDAR and vision-based systems are increasingly paired with machine learning for obstacle
detection. For instance, Pérez et al. (2018) developed a stereoscopic vision system for collision avoidance,
though computational constraints remain a barrier .
 Battery and Security: Emerging research addresses battery management, anti-theft mechanisms, and secure
swarm coordination. Block-chain technology shows promise for secure communication in drone swarms .
 Autopilot systems are tailored to mission-specific needs: Aerial Mapping: UAVs equipped with RTK-GPS and
photogrammetry tools rely on autopilots for precise trajectory control. For example, Backes et al. (2019)
optimized flood hazard modelling using high-resolution drone-based 3D maps . Swarm Autonomy: Control of
Autonomous Aerial Vehicles highlights advancements in swarm coordination, emphasizing AI-driven
autonomy and edge computing for real-time decision-making .
 In the design of an autopilot, three subsystems work together. These are the control subsystem, navigation
subsystem and the Guidance subsystem. This ascertained in Mohammed sadraey book, Design of unmanned
aerial systems.(2020).
 In conclusion, the design and implementation of UAV autopilots focuses on balancing computational efficiency
and adaptability. While open-source platforms and fuzzy logic controllers have advanced autonomy,
challenges in communication reliability and swarm coordination persist. Future research should prioritize AI-
driven control, lightweight architectures, and energy-efficient solutions to expand UAV applications in complex
environments.
3.0 Methodology

 Research and definition

 Dynamic modelling of UAV
 Control system design
 Guidance system design
 Navigation system design
 Machine learning organisms for turbulent weather detection and
avoidance
 Integration of the above systems
 Hardware and software implementation
 Testing and Validation
4.0 Project Plan - Gantt Chart
5.0 Budget Estimate
Component Quantity Total price

Microcontroller 1 3000

Servomotors 4 600

GPS tracker 1 300

Sensors(accelerometer 5-6 1000

etc.)
Android phone 1 2000

Total cost 6900

END