OneFish: Experimental Buoyancy & Autonomous Recovery

Simulation is an indispensable tool—especially for synthetic training via sim-to-real workflows—yet we found that many of the true challenges in multi-agent swarm dynamics only surfaced with in-water field testing. Spanning 2019 through 2024, our Longfish research program focused on the empirical, physical exploration of how to construct and run low-cost drone swarms, culminating in the multi-agent OneFish autonomous swarm. The OneFish vehicles serve as our dedicated internal platform for studying coordinated swarm behaviors alongside practical autonomous deployment, recovery, and maintenance routines.

In 2026, we revived the Longfish program to provide real-world hardware targets as we develop and validate our acoustic UUV detection systems.

Blacktip: 1:1 Biomimetic Testbed & Additive Hull Engineering

We developed Blacktip as an experimental platform to investigate biomimicry for underwater hydrodynamics in unmanned underwater vehicles. Modeled at a 1:1 scale on a blacktip reef shark—a morphology chosen for exceptional gliding efficiency and maneuvering—the vehicle allowed us to build directly upon an evolutionary design space refined over millions of years.

That 1:1 scale also allowed us to explore the practical limits of size, shape, weight, and handling for a marine vehicle designed around a single operator. Despite its size, Blacktip could be deployed and recovered by one person, giving us direct experience with practical design constraints that continue to influence how we approach vehicle design today.

Spanning 2017 through 2021, the program also served as our primary testbed for additive manufacturing in marine robotics. By exploring large-format 3D printing to produce custom, low-cost pressure hulls and hydrodynamic fairings rapidly, the project established core structural and fabrication methodologies that directly inform our current hull construction techniques.

Pathfinder: Autonomous Surface Testbed

We built Pathfinder around a simple idea: if you want to understand autonomous boats, you need to put autonomous boats on the water. Simulation is valuable, but it cannot reproduce every practical problem that appears in the field. Pathfinder gave us a low-cost, highly configurable platform for testing our core autonomy stack, sensors, radios, GPS systems, and other onboard hardware under real operating conditions. It was never intended to become a product, but it informed every vehicle design that followed—and remains a useful test platform whenever we need another autonomous boat on the water.

Built after Blacktip, Pathfinder also gave us a way to explore the opposite end of the design problem: how little infrastructure does an autonomous marine vehicle actually need? We used it to experiment with the minimum practical requirements both onboard and ashore, with a particular focus on deployment, operation, testing, and recovery by a single person. Pathfinder became the platform that taught us which systems truly matter in the field—and what is required to make single-operator autonomous vehicle operation practical.