Abstract
A dragonfly captures its prey with a success rate above 95%, a number that humbles every drone engineer who has tried to replicate insect flight in a laboratory. That gap between biological performance and engineered capability is where defense funding tends to concentrate, and it is also where my two senior research projects meet. My technical capstone develops a bio-inspired Dragonfly Unmanned Aerial System (UAS) that uses four independently actuated wings to mimic the maneuverability of real dragonflies, undertaken because conventional micro-drones still cannot match the agility, efficiency, or stability of the insects they try to copy. My STS research paper examines the military-biological research complex, the network of defense agencies, university labs, and biological organisms that turn scientific curiosity about living systems into deployable military technology, undertaken because the funding behind projects like my own capstone made me want to understand the institutional machinery I was participating in. Rather than running on parallel tracks, the two projects examine the same work from inside and outside the lab. Building the drone is what gave me the questions the STS paper tries to answer, and writing the STS paper is what made me understand what the drone actually represents in a much larger system.
Conventional micro air vehicles hit a wall at small scales. Rotary designs lose aerodynamic efficiency dramatically as rotor diameter shrinks, fixed-wing platforms cannot hover, and most flapping-wing prototypes rely on coupled wing motion that fails to reproduce the agility insects achieve in nature. The ICARUS-1 project seeks to address that gap by designing, building, and testing an Unmanned Aerial System with four independently actuated wings, drawing on dragonfly flight mechanics as the biological reference point for a platform capable of hover, forward flight, and rapid directional changes within a sub-250-gram airframe. The methodology spans three integrated subteams. Aerodynamics built kinematic models from real dragonfly wing motion and used computational fluid dynamics simulations on UVA's Rivanna supercomputer to evaluate angle of attack, wing rigidity, and lift performance against weight. Structures designed a carbon fiber and Mylar wing assembly, a gear-train and crankshaft flapping mechanism, and a weight-optimized 3D-printed body. Avionics integrated a Teensy 4.1 microcontroller, custom PCB, IMU feedback, and a control architecture combining PID and Active Disturbance Rejection Control, validated in a custom simulation environment built on Purdue's open-source Flappy runtime before being moved toward physical hardware.
ICARUS-1 reached partial success against its mission requirements. Independent wing actuation was validated, a sustained dynamic flapping frequency of 25 to 27 Hz was achieved across all four wings, motor encoder tracking met its accuracy spec, and the control algorithm demonstrated stable hover and disturbance recovery in simulation. The wing assembly was iteratively reduced from 3.87 grams to 0.55 grams using vacuum-sealed Mylar over a carbon fiber spar, and the project finished at 80.12% of its $3,000 budget. Full untethered flight was not achieved within the cycle. Early attempts at brushless motors with field-oriented control produced current spikes from over-torquing, which forced a redesign around brushed DC motors driven by pulse-width modulation through an MP6550 H-bridge. The remaining bottlenecks identified for future iterations include lift generation measurement at the single-wing scale (where the strain gauge sample rate approaches the Nyquist limit of the flapping frequency), PLA flexion in the test bench, and the gap between simulated and physical controller tuning. The broader conclusion is that the four-independent-wing approach is mechanically and computationally feasible at this scale, but closing the distance to deployed insect-scale flight will require lighter actuators, more sensitive force measurement, and a hardware-in-the-loop tuning pipeline that the current capstone established the foundation for.
My STS paper asks how military funding shapes the trajectory of biological research and what institutional mechanisms convert biological knowledge into defense technologies. The question matters because defense agencies are now among the largest funders of life sciences and engineering research at U.S. universities, which means the directions science takes are increasingly entangled with military priorities in ways that neither scientists nor the public always recognize. To analyze this entanglement without flattening it into a simple story of progress or corruption, I use Actor-Network Theory, drawing on Law and Callon's (1988) study of the TSR-2 military aircraft project as methodological precedent for treating funding agencies, researchers, biological organisms, and resulting technologies as actants whose interactions produce outcomes no single actor controls.
The paper traces the dragonfly-inspired drone pipeline from the CIA's 1970s Insectothopter through DARPA's contemporary biomimetic programs to deployed systems like the Black Hornet nano-drone, and uses DARPA's Insect Allies program as a case where the dual-use dilemma becomes impossible to ignore. Drawing on Wilson Center funding data, GAO reports, and critiques from Reeves et al. (2018), Salloch (2018), and O'Connell (2023), I find that the benefits and tensions of military-funded biological research are not separable outcomes that can be independently maximized or minimized; they emerge from the same network processes of translation and heterogeneous engineering. The implication is that meaningful governance has to be built into these networks structurally, not bolted on afterward, and that researchers like me, working inside Navy-funded labs on bio-inspired engineering, are part of the network whether we acknowledge it or not.
Notes
School of Engineering and Applied Science
Bachelor of Science in Aerospace Engineering
Technical Advisor: Haibo Dong
STS Advisor: Pedro, Francisco
Technical Team Members: Lily Byers, Kathryn Geoffrey, Jafar Mansoor, Justin Matara, Owen McKenney, Andrew Mercer, Carter Nickola, Jeremiah Nubbe, Nicholas Owen, Mark Piatko, Luis Ramos-Garcia, James Scullin, Matthew Sendi, George Zach
