Project ATLAS Hybrid Rocket Engine; AI in Military Systems: Navigating the Ethical Minefield of Autonomous Warfare

Aerospace technologies are developing at a rapid pace and to ensure responsible use and safety of operation, engineers are faced with solving ethical and physical problems. In rocketry there are three different types of engines: solid, liquid and hybrid. These names refer to the physical state of the oxidizer and fuel within the rocket. Hybrid rockets are the technical focus of this capstone, combining solid fuel and a liquid oxidizer. These rockets achieve a high safety standard when compared to solid and liquid, while still maintaining high power. The sociotechnical aspect of this capstone is an ethical study examining autonomous military systems, comparing them to human soldier decision making frameworks, and offering insight into the future. Though not inherently related, both of these topics are important to study for the advancement of aerospace technologies and to ensure the safety of human lives.
Hybrid rocketry is typically carried out by student led teams, developing on a small scale. These projects are relatively safe compared to solid and liquid-propellant platforms, while maintaining adequate power. Despite this, hybrid thrust production is incredibly difficult to model. Our team developed a hybrid rocket motor to make progress toward better characterizing the performance of hybrids, at low cost. Our project was also intended to set the groundwork for future student led groups to develop hybrid motors. The safety and test procedures can easily be carried over into another project at UVA.
The sociotechnical aspect of this capstone explored the ethical concerns brought upon by the advent of autonomy in military defense systems. This topic is important because military systems affect a vast number of soldiers and civilians globally. Autonomous systems are able to act on their own, making decisions with minimal or no human oversight. Without proper considerations, these decision-making models could cause disaster for many. To start the analysis, various decision making frameworks for human soldiers were examined. This laid the groundwork for how human decision makers are trained to make decisions. There is rarely one specific correct course of action, but soldiers are taught what paths to consider. Next, Actor-Network Theory was used to identify the various actors currently involved in the development and use of autonomous systems. Significant insight from this research is that multiple decision making frameworks will be necessary for responsible AI development. Frameworks like ethical triangulation and moral deliberation posed by Deane Baker and Roger Herbert respectively have prospective strengths in that they consider multiple different perspectives. Ian King’s quasi-utilitarian approach was also found to be useful because of its ability to filter biases through special considerations.
Diving into hybrid rocketry has provided valuable experience in hands-on design, fabrication, and testing. The protocols and procedures developed will serve future UVA students as a starting block for future projects. Autonomous defense systems will only become more prevalent in warfare; it is important to continue discussing how these decision makers will be programmed to act. Global leaders must maintain transparency and bolster collaboration to ensure safety for all in an evolving environment.

School of Engineering and Applied Science

Bachelor of Science in Aerospace Engineering

Technical Advisor: Chloe Dedic & Daniel Quinn
STS Advisor: Sean Murray

Technical Team Members: Harshit Dhayal, Ved Thakare, Mannix Green,
Aiden Winfield, Sean Dunn, Dominic Profaci, Thomas DeCanio,
Joshua Bird, Gavin Miller, Taka Suzuki, Darsh Devkar,
Jack Spinnanger, Isaac Tisinger, Silas Agnew, Zach Hinz,
Alexander Gorodchanin, Adis Gorenca, James Dalzell