Team Swarm Shift
Team Swarm Shift
Columbia University
Swarm Drone Shape Formation and Dispersion
The Team
Aarushi Sharma
• BA in Computer Science & Economics
• Experience in data science, ML, full stack development, technology consulting
Judah Wahba
• BA in Information Science
• Experience working with data science, statistics, software, and financial analysis
Kaya Dorogi
• BS in Computer Science
• Knowledge of drones in construction/ renewables. Full stack software dev and due diligence
Shaun Maher
• BA in Applied Mathematics
• Experience in data science, modeling, Python/Scipl Math, Have pitched to VCs/have experience in capital markets
Problem Sponsor
Dr. Corey Bergsrud, Kira Haag - Naval Surface Warfare Center, Crane Division (NSWC Crane)
Original Problem Statement
NSWC Crane division in Indiana needs to increase exploration and understanding of transformable or shape changing unmanned aerial systems (UAS) and unmanned systems (UxS) to apply to existing warfare capabilities and novel future warfare concepts.
Beneficiary Discovery Interviews
30
The Innovation
Hacking for Defense (“H4D”) students are aiding the U.S. Department of Defense in advancing drone swarm capabilities for resilient autonomous operations. The Columbia University student team up for the challenge included Judah Wahba, Aarushi Sharma, Shaun Maher, and Kaya Dorogi, with diverse academic backgrounds including computer science, information science, and mathematics. “Previously, I had the privilege of working on a few different government projects which honestly instilled a sense of patriotic duty, and I believed a student at Columbia could make a difference. H4D seemed like a perfect way to continue that impact in my final year," said Maher. While Dorogi had general knowledge and a passion surrounding commercial drones, none of her team members had technical experience with programming drones or military-drone use-cases.
Initially, the student team was entrusted by the Naval Surface Warfare Center, Crane Division (NSWC Crane) with the challenge of expanding shapeshifting drone technology for battlefield use by the Navy. They aimed to enhance their knowledge and research on transformable or shape-changing unmanned aerial systems (UAS) and unmanned systems (UxS) in general. This effort is intended to improve current warfare capabilities and develop innovative future warfare concepts. "As we explored the challenge, we had the option of focusing on developing the architecture of a drone capable of changing its shape mid-air. However, we chose to investigate different swarms of drones instead,” Wahba explains. “Our goal was to understand how, without altering the structure of individual drones, we could manipulate the shape of a swarm mid-air to enhance deception capabilities for the Navy.” With the freedom to explore various aspects of drone technology, the students decided to embrace their coding strengths and pivot away from traditional hardware solutions. “Essentially, our approach revolved around researching how our skill set could be best applied to the project."
The students conducted 30 interviews with experts across the defense space, including the National Security Innovation Network (NSIN), Defense Advanced Research Projects Agency (DARPA), and NSWC Crane. One of their key findings in the interviews was that there were significant operational limitations with having humans directly controlling individual drones in contested environments. "A big part of our learning was emphasizing a thorough upfront discovery process rather than jumping straight to building technology," noted Wahbah. Their discovery "allowed us to really understand the complexities of operating drones in military environments."
The team landed on a minimum viable product demonstration showing the autonomous leader election capability, adapting to the team’s software background. "We created an interface that follows a leader-follower approach in the drone swarm... creating a self-managed swarm of drones," said Sharma. Their presented approach utilized the Paxos algorithm, allowing the drones to autonomously elect a "leader" that could dynamically change the swarm's flight formation and behavior. Maher explains, "Paxos basically comes up with a protocol where you can deal with leader election for drones or for any kind of distributed system, and then deal with reliably propagating commands to the rest of the nodes in the network." If the leader drone was disabled, the other drones would seamlessly vote and elect a new leader, ensuring resilient swarm operations. The Navy problem sponsor was impressed with the novel approach and the team's adaptability.
The Impact of Team Swarm Shift’s H4D Experience
Team Swarm Shift’s innovative drone swarm solution exemplifies the transformative impact students can have on impacting real-world national security priorities through cross-disciplinary collaboration and embracing an iterative problem-solving mentality honed through H4D's methodology. Wahab affirms, "The impact of our project, especially being able to contribute to something as significant as military technology, really emphasized the practical applications of our studies."
Maher underscored the prevalent urge within tech-related spheres to prioritize speed over precision when tackling software and hardware challenges. What struck him as distinct about their experience with H4D was the deliberate shift in tempo—a conscious decision to pause and pose the right questions. Maher explained, “H4D emphasized pausing to ask the right questions, enabling more efficient planning and solution-building to tackle the most crucial problems effectively.” “As engineers, we don't typically envision all the potential use cases for the technologies we learn,” said Aarushi.
“Hacking for Defense exposed us to a whole new domain of opportunities around national security challenges,” said Aarushi. The experience expanded their interests in working on impactful problems in big tech, defense, finance, entrepreneurship, or consulting after graduation. “Working together to solve the ambiguity in our problem left me with similar problem-solving tactics I'll utilize after I graduate, in consulting,” said Dorogi.