- FARAMIR Project
For my senior design project, I collaborated with five other engineering students on FARAMIR (Free-space Airborne Radio Antenna Measurement Integrated Receiver), an industry-sponsored project sponsored by General Atomics. The goal was to create an autonomous drone-based solution for low-frequency antenna testing as a cost-effective alternative to traditional anechoic chambers. As the lead mechanical engineer, I focused on designing and fabricating the payload housing, antenna mounts, and non-conductive landing gear, guiding the mechanical aspects of the project from initial concepts through prototyping and final integration.
This website page outlines our full design process and outcomes. It begins with the problem statement and systems-level design, then breaks down the electrical and mechanical engineering work into the payload, antenna systems, and alternative landing gear. Additional sections cover manufacturing, assembly, and final results.
Problem Statement
Traditional antenna testing relies on anechoic chambers to isolate signals and measure radiation patterns with high accuracy. However, as antenna sizes increase and operating frequencies decrease, the chambers required for testing can become massive. Radio frequency engineers at General Atomics even compared the size of the needed chambers to a football field. They are also expensive to build and operate, making them impractical for rapid prototyping and design companies such as General Atomics.
With larger low-frequency antennas becoming increasingly common in the aerospace and defense industries, the FARAMIR project addresses this challenge by developing a drone-based testing platform that is portable, scalable, and cost-effective. By carrying a lightweight payload with integrated transmission hardware, the drone can autonomously fly hundreds of feet away in precise flight paths around an antenna under test, collecting the data needed to accurately characterize its performance. This approach not only reduces costs dramatically but also enables testing in open environments, providing flexibility and accessibility that traditional methods cannot achieve.
Figure 1. World’s largest anechoic chamber, spanning 264 × 250 × 70 ft with a 4.62 million cubic ft volume. (The Benefield Anechoic Facility at Edwards Air Force Base)
Figure 2. General Atomics’ concept for a drone-based antenna testing platform, which our team developed into the FARAMIR system as an alternative to large anechoic chambers.
System Level Design
The FARAMIR system is built around two integrated subsystems: an airborne payload carried by an industrial-grade drone and a ground-based control and antenna measurement station. At a high level, the drone subsystem provides mobility and transmission capability, while the ground system handles data collection, processing, and supplies power to the antenna under test (AUT).
Onboard the drone, the payload houses all transmission hardware, including a programmable signal generator, power amplifier, and antenna mounting structures. Commands from the ground are relayed to this payload through a wireless link, allowing the drone to transmit test frequencies while maintaining precise autonomous flight paths. Mechanical design choices focus on weight reduction and non-conductive materials so that the housing, landing gear, and antenna fixtures do not distort radiation measurements.
The ground system includes the AUT, a spectrum analyzer with low-noise amplification, and a control interface that synchronizes frequency data with the drone’s GPS and telemetry. By pairing the signal strength recorded on the ground with the drone’s position in flight, the system reconstructs gain patterns for the AUT.
This system-level approach, splitting functionality between air and ground, ensures that the drone operates as a controlled, mobile transmitter while all high-fidelity measurement and analysis stay safely on the ground. The design emphasizes flexibility, allowing components to be swapped or upgraded without redesigning the entire system.
Figure 3. System-level block diagram showing the airborne transmit subsystem housed on the drone and the ground receive subsystem. Arrow and block colors correspond to the key table (right)
Electrical Components
Disclaimer: The exact models and manufacturers of the electrical components used in the FARAMIR project cannot be disclosed, as they are proprietary to General Atomics and considered sensitive information. The following section describes the function and integration of each component type within the system rather than specific component details.​​
Airborne Transmit Subsystem
All payload electronics are mounted to the drone and powered through its onboard battery. Power regulation is handled by multiple buck regulators, which step down the high-voltage supply from the drone into stable low-voltage lines suitable for the onboard microcontroller and RF components. At the core of the payload is a microcontroller that manages communication links with the ground, interprets flight commands, and directs the RF signal chain. This chain consists of a programmable signal generator that produces the test frequency, an RF power amplifier that boosts the output, and a swappable transmit antenna that radiates the signal toward the antenna under test.
The airborne payload begins with a programmable signal generator, which produces the test frequencies within the required operating band. These signals are then passed to a power amplifier, boosting the output to ensure they can be measured accurately at the ground station. Finally, the amplified signal is delivered to a swappable transmit antenna mounted beneath the drone. The antenna design is modular, allowing vertical or horizontal polarization configurations and enabling testing across a variety of frequency ranges without redesigning the payload.
Flight control is managed by an autopilot system integrated with GPS, ensuring the drone maintains precise flight paths around the antenna under test. The payload electronics are designed to accept commands from the ground station, enabling real-time adjustments to transmission frequency during flight. At the same time, the drone remains compatible with manual pilot override through a dedicated control receiver, ensuring safe operation and recovery in the event of autonomous system errors.
Figure 4. Internal view of the payload enclosure showing the microcontroller board, voltage regulation, and signal generator hardware.
Figure 5. Externally mounted power amplifier attached to the payload housing, drawing power from the drone battery.
Figure 6. Swappable transmit antenna mounted beneath the payload housing during outdoor testing.
Ground Receive Subsystem
Power is supplied externally, through a portable generator. The ground system computer serves as the central hub, logging flight data and coordinating antenna measurements.
A dedicated low-noise amplifier (LNA) conditions signals received from the AUT before feeding them into a spectrum analyzer, which captures frequency, amplitude, and timing information. This data is synchronized with GPS and drone telemetry so that each measurement corresponds to a precise drone position and angle relative to the AUT. A dual-channel communication setup ensures redundancy: one link controls flight operations, while another manages payload transmissions.
Mechanical Components
The mechanical design of FARAMIR was critical to the success of the project, as every electrical subsystem and antenna relied on a stable, lightweight, and non-interfering structural platform.
Over the nine-month project, multiple iterations of housings and antenna mounts, and legs were designed, fabricated, tested, and refined.
Payload Housing
The payload housing served as the central enclosure for the signal generator, power amplifier, voltage regulators, and communication hardware. One of the primary mechanical challenges was the tight spatial constraints imposed by the Aurelia X6 drone frame. Two parallel carbon beams ran underneath the drone body, restricting the maximum housing width to 6.04 inches. This dimension directly influenced the CAD geometry, forcing a rectangular footprint. The RF team at General Atomics also required that the housing height remain as short as possible, since the vertical antenna mounted beneath the payload was limited by the clearance of the enclosure. To maximize the effective length of this antenna for low-frequency performance, the payload design targeted a profile of roughly 2.5 inches, pushing the team to reduce unused internal volume and externally mount larger components such as the power amplifier.
Prototype Iterations:
-
Prototype 1–3 (PLA, 20–30% infill): Early housings were printed in PLA for speed and cost efficiency. While structurally sound, these designs were heavier (~3.6 lb) because we had little experience setting optimal infill percentages and overbuilt the walls. PLA also provided limited thermal resistance, and in these early versions the power amplifier was internally enclosed, which consumed unnecessary space.
-
Prototype 4–5 (PLA, reduced wall thickness): The walls were thinned in an effort to cut weight, but concerns remained. The General Atomics team flagged the internal placement of the power amplifier as inefficient and added needless weight to the payload. These prototypes highlighted the need to mount the amplifier externally and to rethink the housing’s overall geometry.
-
Prototype 5-7 (PETG, 10–15% infill): Switching to PETG improved thermal resistance and strength, allowing thinner walls without compromising rigidity. Externalizing the power amplifier reduced internal clutter and lowered the enclosure’s profile. By the seventh iteration, the team had optimized the design to 10% infill PETG, yielding a housing that weighed just 0.612 lb including the lid, clamps, and threaded brass inserts.
Figure 7. Early SolidWorks model of the payload housing with the power amplifier enclosed internally; mass measured at 3.60 lb with clamps and lid.
Final Design: The final housing (6.04 × 6.81 × 2.50 in) met all dimensional constraints, secured electronics with brass threaded inserts, and was clamped to the drone using four PETG brackets. This design strikes a balance between lightweight construction, thermal stability, and ease of assembly, while keeping the footprint compact enough to fit beneath the industrial-grade drone’s restricted frame geometry.
Figure 8. Final CAD model and 3D printed payload housing with threaded inserts.
Antenna Clamps and Mounting System
To fully characterize the antenna under test (AUT), the system needed to support measurements in both linear polarization and magnetic field coupling configurations. This requirement drove the design of two distinct mounting systems: a vertical monopole-style antenna and a large loop antenna.
-
Vertical Antenna Mount: Directly integrated into the bottom of the payload housing with steel 6-32 screws. By keeping the payload housing as short as possible, the vertical antenna length was maximized, improving sensitivity at the lowest frequencies tested.
-
Loop Antenna Mount: Mounting the loop required a rigid, non-conductive fixture to avoid distorting the radiation pattern. A PETG clamping system was designed, consisting of perimeter clamps, a center support, and plastic fasteners to eliminate conductive paths.​
​
Prototype Development
The loop antenna fixtures were designed to secure the top and bottom of the semi-rigid loops. Three different loop sizes were requested by the RF team, and all were accounted for in the design. A cross-shaped fixture connecting to the alternative landing gear was chosen to minimize weight while maximizing structural integrity.
The integration of the clamp system into the alternative landing gear allowed the loop diameter to be expanded to the maximum practical size, directly improving sensitivity at low frequencies.
Figure 9. Vertical antenna mounting design (left) and the physical showcase (right).
Figure 10. Original concept design for loop antenna mounting created in Blender.
Figure 11. Final CAD designs for alternative landing gear and loop antenna clamps, combining 3D-printed fixtures with PVC tubing.
Alternative Legs & Drop Impact Analysis
The original Aurelia X6 landing legs were made of carbon fiber, lightweight but conductive, which created interference with radiation patterns. This required a pivot to custom, non-conductive landing gear fabricated from PVC tubing and fiberglass rods.
Design
-
Materials & sizes: PVC base and fiberglass struts, both 1.05″ OD × 0.842″ ID.
-
Geometry: Square PVC base 15″ × 15″; two 17″ fiberglass struts set at 70° to the PVC base (20° to vertical) into the PVC joints and the drone’s original leg-sockets.
-
Dual purpose: Provides landing support and a stable, non-conductive platform for the loop antennas.​​
Figure 12. Schematic of the alternative landing gear design using PVC and fiberglass rods to provide a non-conductive base for antenna mounting.
Structural Analysis
Hand calculations (6-ft drop, total weight = 15 lbf)
Using energy methods and cantilever beam theory, the two fiberglass struts dominate the vertical compliance:
-
Combined vertical stiffness: ~283 lbf/in
-
Predicted peak deflection: ~2.76 in
-
Peak impact load: ~783 lbf total (~392 lbf/strut)
-
Peak bending stress (per strut): ~100 ksi
-
Factor of Safety (fiberglass allowable 150 ksi): ~1.50​​
Figure 13. Drop impact analysis of the alternative legs, using the hand calculation method.
Abaqus Analysis
Cleaned SolidWorks geometry was imported to Abaqus. A 6-ft drop was modeled with gravity and initial velocity v= sqrt(2gh). The ground was a rigid analytical surface with hard contact (normal) and frictionless tangential behavior. Meshing used C3D10 tetrahedra (2–3 mm at PVC/fiberglass interfaces, 8–10 mm elsewhere). A mesh convergence study (<5% change in max stress/strain energy) was completed, and simulation results matched the hand calculations within ±10%, confirming model fidelity.​
​
Outcome: The non-conductive landing gear resolves EM interference while meeting structural targets for the 15 lbf, 6-ft drop case with FoS ≈ 1.62 on the fiberglass struts.
-
Max deflection: ~2.64 in
-
Peak impact load: ~734.3 lbf
-
Fiberglass max stress: ~92.59
-
FoS (150 ksi allowable): ~1.62
Ground Subsystem
The ground housing was initially considered for 3D printing. Early CAD concepts included ventilation slots, lid pillars, and threaded bosses for mounting.
​
Specifications
-
Dimensions: 9 × 12 × 6 in
-
Material: Commercial-grade plastic (purchased enclosure)
-
Contents: Spectrum analyzer, low-noise amplifier, portable power supply, and fiber-optic converter
​
Decision Rationale
Ultimately, the team chose to purchase a plastic box from Home Depot instead of 3D printing a large enclosure. This choice was driven by:
-
Time efficiency: Printing would have required buying a roll of PETG and over 40 hours of print time.
-
Durability: Off-the-shelf box provided ruggedness and handles for portability.
-
Flexibility: Internal 3D printed mounts were still used to secure components, combining practicality with customization.
This solution was not a compromise under deadline pressure, but a deliberate decision balancing cost, strength, time, and taking the simplest path to solving an engineering problem.
Figure 14. Original CAD design for the ground housing. The internal mounts for the spectrum analyzer and low-noise amplifier were 3D printed, while the enclosure itself was replaced with a cost-effective plastic box from Home Depot.
Fabrication and Assembly
Electronics Fabrication
Soldering was the primary fabrication method. Components such as the microcontroller and LoRa communication module were hand-soldered onto protoboards. Flux-core solder was used to ensure clean joints and long-term durability. Wire gauges were chosen based on current demand: thin wires for low-power data lines and thicker leads for the amplifier, which required higher current handling.
Once the major connections were complete, the boards were tested individually for continuity and voltage stability before being integrated into the housing. This step was critical for catching errors early, since troubleshooting after installation would have required partial disassembly. Caution was also taken with the power amplifier wiring, as incorrect connections could have damaged both the amplifier and the generator. Finally, connectors were added to the outputs, ensuring that the payload could be quickly attached to the antenna mounts, secured inside the housing, and interfaced with the drone’s power system.
Mechanical Assembly
Mechanical assembly focused on combining the fabricated electronics with the custom housings, clamps, and landing gear. The payload housing, manufactured from PETG using 3D printing, was fitted with heat-set brass threaded inserts to allow secure fastening of both the electronic boards and external clamps. These inserts ensured that repeated assembly and disassembly would not strip the plastic.
Electronics were first screwed within the housing, each component assigned a dedicated location with its own bosses and threaded inserts to balance weight distribution and maintain airflow. The signal generator and voltage regulators were placed internally, while the power amplifier was mounted externally on the housing wall to reduce thermal buildup and free up internal space. Cable routing was carefully planned in the design process.
The housing was attached to the underside of the drone frame using lightweight clamps designed to withstand vibration while keeping the enclosure aligned with the drone’s center of gravity. The vertical antenna was mounted directly to the bottom of the housing, while the loop antenna used a clamping system integrated into the alternative landing gear, allowing the loop to be as large as possible: an important factor for low-frequency antenna testing. Each mounting system was checked for rigidity to prevent unintended shifts that could skew results.
The payload housing, antennas, and alternative landing gear were assembled onto the drone in stages, with checks at each step to verify clearance, alignment, and weight balance. By the end of this process, the drone was fully assembled with its electronics secured, antennas mounted, and amplifier installed, forming a complete system ready for integration and flight testing.
Results
The prototype successfully demonstrated the feasibility of a drone-based low-frequency antenna testing system as a cost-effective alternative to traditional anechoic chambers. Through subsystem testing and full-system integration, our team validated the mechanical durability of the payload housing, the strength of the non-conductive landing gear, and the stability of antenna mounting fixtures during flight.
Flight tests confirmed that the drone could execute autonomous circular paths around an antenna under test while transmitting at programmed frequencies. Ground station measurements showed that received signals aligned with expected trends of characterizing the antenna's performance.​
Accuracy Comparison
When benchmarked against data collected in General Atomics’ anechoic chamber, the drone-based measurements achieved an accuracy within ±15% of expected values. With further post-processing and testing in a more secluded environment, this margin of error is expected to be reduced to as low as ±5%. For this project, however, testing was conducted on GA-owned property in an open baseball field - an environment where external factors such as lighting structures and nearby infrastructure introduced unwanted signal reflections and interference. Despite these challenges, the prototype’s accuracy demonstrated strong potential for refinement.
Due to the nature of this project, detailed numerical results, radiation patterns, and raw performance data cannot be shown publicly, as this information is considered proprietary.
Even with these restrictions, the overall outcome of the project was a functioning prototype that met core design requirements and highlighted a clear path forward. Future improvements, such as more robust communication links, improved power conditioning, and expanded multi-frequency capabilities, can build directly on this foundation. Engineers at General Atomics intend to deploy FARAMIR for future antenna testing, validating our work as a meaningful contribution to the aerospace and defense company.
Video 1. Live demonstration of the FARAMIR drone operating.
Video 2. Demonstration of the LoRa communication link, showing feedback from the payload regarding the AUT while the FARAMIR drone follows its autonomous flight path.
Meet the Team
FARAMIR was brought to life by a six-person team of senior engineering students (three mechanical and three electrical) at the University of San Diego, working in partnership with General Atomics’ Radio Frequency team.
Charles Gorey (Mechanical Engineer)
Led the design of the vertical and loop antenna mounting systems. Manufactured alternative landing gear and assisted with the assembly of all mechanical components onto the drone. Focused on structural analysis, hand calculations, and the integration of non-conductive materials.
Charles Gorey completed a summer internship with a stealth startup in the space and satellite industry and will be returning there in January 2026 after finishing his final semester at USD.
Patrick McDermott (Electrical Engineer)
Electrical engineering lead who worked extensively on the communication aspects of electrical and mechanical system integration. Led RF design, including link budget development, programming and testing of the signal generator, and integration of the power amplifier to meet test requirements.
Patrick McDermott is completing his final semester at USD and plans to attend graduate school in 2026 to continue studying electrical engineering.
Luca Sacchetto (Mechanical Engineer)
Led the mechanical design and integration of all systems to fit compactly beneath the drone’s body. Designed the payload housing using SolidWorks and conducted stress analysis on the alternative legs using Abaqus. Contributed to the manufacturing and assembly of all mechanical components.
Luca Sacchetto is currently pursuing his master’s degree in Aerospace and Mechanical Engineering at the University of Southern California, focusing on computational fluid and solid mechanics.
Jonathan Miller (Electrical Engineer)
Embedded systems and microcontroller programming, integrating wireless communication, signal generation, and ground system data logging into a unified control system.
Jonathan Miller is now working as a Project Engineer at Precision Partners Industrial Development Group (PPIDG).
Adam Walter (Flight Engineer)
Certified FAA Part 107 drone pilot responsible for flight operations and Mission Planner programming. Assisted with manufacturing, final mechanical assembly, and antenna installation.
Adam Walter is completing his final semester at USD and plans to pursue a career in the aerospace industry after graduating in December 2025.
Fernando Huerta (Electrical Engineer)
Designed power distribution for both drone and ground subsystems, and soldered and assembled payload electronics. Built a prototype drone to validate proof of concept early in the first semester.
Like Adam, Fernando Huerta is completing his final semester at USD and plans to pursue a career in the aerospace industry after graduating in December 2025.
Kanan Schmid (Project Manager)
Project lead from General Atomics, providing technical guidance, resources, and mentorship throughout the nine-month capstone. Defined requirements, evaluated design decisions, and led weekly meetings with both the USD and GA teams.
Kanan Schmid graduated from the University of California, Los Angeles (UCLA) in 2021 with a bachelor’s and master’s degree in Electrical Engineering, specializing in RF engineering and electromagnetism. He has since been working at General Atomics and is now a Senior Engineer.
