Self-healing Battlefield Communications

Final In-Class Presentation

In war, the most dangerous sound is silence — not from the enemy, but from your own comms.

Team Omnicomm entered Stanford’s Hacking for Defense course with one mandate: make battlefield communications resilient. Not just more secure or better encrypted. Actually resilient. Their problem sponsor, the 18th Airborne Corps, made it plain. In combat, most forward teams juggle one or two connections at best. Failover is manual. Jamming is common. The routers blink red. And when that happens, mission control does not just lag. It disappears.

By the end of the course, Omnicomm had field-tested a prototype router that bonded commercial and military networks, autonomously rerouted around failure, and delivered uptime in motion. They were not just patching the gaps in today’s tactical architecture. They were designing it to survive tomorrow.

What Was at Stake

In today’s fight, the difference between success and silence is measured in seconds. And most military communications systems fail that test. They rely on single transport layers, siloed protocols, and operator-dependent switching. In the fog of combat, warfighters were juggling radios and networks that could not talk to each other — or worse, failed to talk at all.

Omnicomm reframed the mission. What the force needed was not just faster throughput or hardened encryption. It was software-defined, modular networking that worked with gloves on, stress high, and no tech in sight. Systems that adapted in real time, across Starlink, LTE, ROIP, and mesh networks, without asking the operator to lift a finger.

The goal was simple. Plug in, power up, and keep moving, even when everything else breaks.

The Discovery Loop

The team ran 137 interviews across the course, starting with signal officers and NetOps staff, and quickly moving to forward operators in Special Operations, rapid response units, and contested COC environments. Their interviews surfaced a consistent failure point. Existing architectures were stovepiped and fragile. Transport layers did not talk to each other until it was too late. Units were relying on apps like Signal to backfill broken systems.

The technical ask was not impossible. It just required someone to build it for the user, not the specs. So the team got out of the building — and into the field.

At SOF Week in Tampa, they brought a functional prototype made from off-the-shelf parts. The prototype was not perfect. It broke in transit. But it worked. And the feedback confirmed what operators had been saying for years: plug-and-play autonomy is not a luxury. It is a mission need.

MVP That Works Like Warfighters Do

Omnicomm’s first MVP was a low-cost, modular router with seamless failover and bandwidth bonding. It was designed to ingest whatever connections were available — commercial or military — and deliver a single, stable pipe to the operator.

The interface was built to be invisible. One-button activation. No NetOps reconfigurations. No protocol juggling. Just uptime.

Their system did not require new networks. It integrated the ones already in use — faster, smarter, and with real-time intelligence on link health and automatic rerouting.

Operators of Another Kind

• Luke Virsik, software engineer, served as both technical lead and product manager, translating field needs into interface logic and prototype specs.

• Andrew Paulmeno and Samuel Montagut, both hardware engineers, built the physical unit, optimized network adapters, and handled bandwidth load balancing across multiple sources.

• Charlie Gordon, software engineer, focused on interoperability and UI simplicity, ensuring seamless operation under pressure.

• Facundo Cabrera-Booman, PhD, brought advanced systems thinking to encryption redundancy and performance mapping.

• Althea Hudson, software engineer, led backend software development for routing logic and tested reliability under simulated denial conditions.

They were mentored by CW4 Jason Horton of the 18th Airborne Corps and supported by Adam Waters and the broader CMP instructional team.

What Comes Next

Omnicomm left the course with a working prototype, real-world feedback from SOF Week, and a clear roadmap. The team is exploring transition pathways through DIU, SOCOM S&T, and upcoming TES 25-3 exercises. They have raised early-stage capital through their spinout, Hoplynk, and are pursuing a dual-use roadmap to reach TRL 6 and place commercial units in production.

They are not building just a better router. They are building the new default.

They did not just ask how to improve tactical comms. They built a system smart enough to survive them.

Read more on the Stanford H4D Substack