Detect
Cyclostationary feature detection identifies gossip protocol timing signatures without prior knowledge of drone type or manufacturer.
Early Development — Taiwan 2026
Detects and disrupts autonomous drone swarms without jamming
SwarmSense is a software-defined counter-swarm detection platform for critical infrastructure and defense.
The economics have inverted
$3,000,000
Patriot missile interceptor
$500
FPV autonomous attack drone
$0 marginal
SwarmSense protocol disruption
- Jamming is obsolete — modern drone swarms are designed to operate without any operator control link.
- Signature matching fails — autonomous swarms built for warfare have no commercial signatures to match.
- Kinetic intercept cannot scale — one Patriot battery cannot economically counter a 200-drone swarm.
Attack the coordination layer, not individual drones
Autonomous swarms coordinate using gossip protocols. The RF timing signature of this coordination is detectable and disruptable even below the conventional noise floor.
Analyze
Graph-theoretic inference maps swarm communication topology and identifies high-degree connector nodes for targeted disruption.
Disrupt
Surgical narrowband signal injection targeting consensus timing. Legally distinct from jamming. Zero marginal cost per engagement.
Taiwan is the primary deployment market
Threat context
- Taiwan's air defense is optimized for ballistic and cruise missiles — not low-altitude autonomous swarm attacks on critical infrastructure.
- Taiwan's ports, airports, science parks, and submarine cable landing stations have no systematic autonomous swarm detection capability.
Opportunity
- AppWorks AW#33 defense tech focus provides accelerator access and NCSIST relationships.
- Tier 1 detection-only deployments require no special regulatory authorization — commercial revenue without defense procurement.
- Deploy Taiwan first as proof-of-concept. Use that deployment to raise international capital.
Global opportunities
Y Combinator's Summer 2026 Counter-Swarm Defense request for startups is a useful market signal: tier-one accelerators are explicitly backing companies that treat drone defense as a real-time distributed systems problem—not only a weapons problem. Partner Tyler Bosmeny frames the winning pattern as "Cloudflare for the physical world": infrastructure that can absorb asymmetric swarm economics (e.g. a low-cost drone against a high-value asset) and scale to dozens of simultaneous engagements, where single-target workflows break down.
Layers YC highlights in the stack
High-capacity interceptors
Mass-producible kinetic or non-kinetic defeat so defenses don’t “run out of bullets” against swarms.
SwarmSense
Sensor fusion software
Radar, RF, acoustic, and more fused into one real-time picture across hundreds of tracks.
Non-kinetic defenses
EW, aerosols, streamers, and other non-explosive ways to degrade or disable drone electronics.
SwarmSense turns the swarm’s own radio chatter into a live coordination map—no aircraft catalog and no pilot link required.
That framing aligns with what operators already feel: cost parity must flip, engagement must be increasingly AI-native, and dual-use protection of ports, airports, and energy sites sits beside military demand.
"Drone defense is looking less like operating a weapon and more like running a real-time distributed system."
The Team
Paul Yang
FounderSignal processing engineer with an MS in signal processing and 20 years of software engineering experience. Building the first protocol-layer counter-swarm detection system for Taiwan's critical infrastructure.
Seeking a technical co-founder with RF, EW, or drone systems experience. mjpablo23@gmail.com
Work with us
Researchers & collaborators
We are seeking research collaboration with signal processing faculty and graduate students in Taiwan. If your work touches cyclostationary analysis, cognitive radio, or RF machine learning, we would value your perspective.
mjpablo23@gmail.comPotential customers
If you are responsible for security at a Taiwan port, airport, science park, or energy facility and autonomous drone threats are on your planning horizon, we want to talk.
mjpablo23@gmail.com