We believe the living water deserves continuous, transparent observation — and we are building the workforce to provide it.
A traditional observatory is not a telescope. It is a long-running institution that points telescopes at the sky, records what it sees with care, and publishes the record so that others can build on it. The instruments change. The practice of observation does not.
Underwater observation deserves the same posture, and has rarely been given it. The vehicles that descend into reefs, kelp forests, lakes, and littoral zones are increasingly capable. The institutional practice around them — the way the resulting facts are documented, signed, owned, and shared — has not kept pace. We are building both.
Hardware moats are increasingly fragile. Component supply chains, fabrication techniques, and reference designs are converging across the industry, and the conservation customers we serve cannot afford lock-in regardless. Our vehicles are built on published designs, documented bills of materials, and open firmware. The artifact is not the moat.
Every observation an Observatory vehicle produces is signed at the moment of capture — vehicle identity, pod identity, firmware hash, calibration state, timestamp. The signed record travels with the data forever. The customer owns it. Two decades from now, anyone holding the public key can verify that a 2027 recording is what the recording claims to be. Provenance is infrastructure, not metadata.
We do not sell vehicles. We sell deployments — reef-weeks, lake-seasons, harbor-quarters. A customer pays for continuous observation of a place, performed by a coordinated fleet operated by people who have done this before, with the data flowing into a canonical record that the customer owns and the public can verify. The vehicle is a means. The continuous record is the end.
These three commitments are load-bearing. They constrain the company more than they free it, which is the point. An observatory that is not open, not sovereign, and not labor-shaped is a vendor with a marketing department.
The Observatory is the public face of every deployment we run. When a fleet is in the water, this page shows what the fleet is doing in close-to-real time — vehicle positions, recent observations, signed records you can verify in your browser. When no fleet is deployed, it shows the most recent observations and the full archive of every deployment the company has ever produced.
Below is the structural form of an Observatory record, populated with synthetic sample data so that the shape is legible before the first deployment fills it.
A low-frequency rumble detected at 14m depth, matched against a candidate sturgeon vocalization profile with moderate confidence. In a live record, this card would link to the verified audio clip and the full detection trace.
The form a deployment summary takes once the fleet has surfaced and the records have been signed, verified, and archived. The same record persists for the life of the company.
Pilot a free-swimmer through six real water bodies under live conditions. Turn on live data to seed each site from active NOAA and USGS stations.
The inland sea is the finest freshwater preserve of shipwrecks on Earth. Its cold water keeps hulls, timber, and painted lettering intact for a century and more. Every wreck is a documented fact — a known name, a known date, a documentation history that travels with it. That makes Lake Superior the truest possible subject for a provenance-first observatory, and our founding emphasis.
Identity, launch, final voyage, the storm that took her. Each record is source-stamped to the archive it came from.
Depth, orientation, state of preservation — and the live conditions over the site right now, drawn from the nearest station.
Protection status, monitoring needs, threats — the wreck as a fragile, non-renewable cultural resource.
A real workforce has roles. Our deployments combine tethered tenders — anchored, high-power, comms-rich — with battery-powered free-swimmers that range, listen, and dock periodically to the tender for charge and data offload.
Every vehicle runs the same three-tier compute architecture: a deterministic reflex layer that handles thrusters, leak detection, and emergency surface; a coordination layer that runs the mission state machine and acoustic communications; and a cognition layer — Raspberry Pi 5 for acoustic work, Jetson-class for vision — that handles the bioacoustic matching, anomaly detection, and the natural-language interface to the operator.
The vehicles are built on Blue Robotics components and our own pressure-housing designs. The CAD, BOMs, and firmware are published. Anyone can build one. Almost nobody can run a fleet of them well — and the work of running them well is what we sell.
Four-channel hydrophone array, Pi 5 cognition, closed-loop signal matching. Direct inheritance from our ORCA bioacoustic platform.
Stereo camera with onboard lighting and Jetson-class classification. Reef, kelp, lake-bottom mapping with verified provenance per frame.
CTD, dissolved oxygen, pH, turbidity, fluorometry. The unglamorous baseline every customer asks for first.
Small manipulator with water and sediment collection. The slow, expensive one. Scheduled for v2.
Observatory is in its earliest phase. There is nothing here yet because there is nothing yet to report. The first entries will document our freshwater bench work, the construction of the first listener pod, and the path to the first field deployment.
Observatory was founded in 2026 to build the practice of underwater observation that the moment requires: continuous, transparent, sovereign, and open at the substrate. We are headquartered in Minneapolis with active institutional ties in Hawaiʻi.
The company emerges from a longer body of work on federated data infrastructure, provenance-forward systems architecture, and bioacoustic signal analysis — work that has already produced research nodes in Hawaiian botany, archaeological survey, and museum publication. Observatory is the marine and freshwater node of that federation. The three commitments on the Approach page are not aspirational; they are the way the rest of the federation already operates.
We are starting in freshwater because freshwater is where we live. Minnesota's ten thousand lakes, the Boundary Waters, and Lake Superior are not a detour from a saltwater story — they are the first chapter of an observatory whose subject is the living water in every form it takes.
Observatory will hire its first team in the second half of 2026, after the first freshwater deployment has produced a record that can stand on its own. The roles below are the shape of that team. If any of them describe work you have spent years preparing for, write to us now.