Ocean robots shouldn’t need a ship to recharge.
Cove is a floating depot that turns wave motion into electricity and lets autonomous underwater vehicles dock, recharge, and go back to work.

Fig. 2 Animated schematic, not to scale
Panel angle follows wave slope; pivot brightness is generator activity. The bar is stored charge; the hull beneath is a docked AUV.
The ship is the bottleneck.
Autonomous underwater vehicles (AUVs) survey pipelines, map the seafloor, and monitor the ocean. They run on batteries that last a day or two. Then a crewed ship has to find them, haul them out, and recharge them.
- Endurance
- 1 to 2 days
When the battery is gone, the mission ends whether the job is done or not.
- Energy on board
- 4 to 5 kWh
What a survey-class AUV carries. It has to cover propulsion, sensors, and the trip back.
- Dominant cost
- The ship
Vessel time, crew, and fuel dominate the cost of running AUVs. The vehicle is bought once; the ship is paid for on every mission.
- Charging at sea
- None unattended
There is no self-powered infrastructure for a vehicle to recharge unattended away from a cable or a ship. Every operator solves it the same expensive way.
How Cove works.
A wide floating frame with rows of hinged panels. Waves rock the panels; the panels turn generators; the generators fill a battery bank; the battery charges docked vehicles.
Fig. 3Cross-section of one bay. Schematic, not to scale.
Waves rock the panels
Each bay holds a curved panel on a pivot. Passing waves swing it back and forth.
Panels drive generators
The pivot turns a gearbox and generator. Many small units, not one big one, so the platform keeps working if one fails.
Power goes to a battery bank
Energy is stored on the platform, so charging does not depend on the sea state at that moment.
AUVs dock and recharge
Vehicles dock in a sheltered bay, charge through an inductive connector, offload data, and leave on a full charge.
The wave-energy physics here is well established; what is new is packaging it as a charging depot for robots. Our energy model runs on real Gulf of Mexico buoy data; its outputs are screening estimates, not measurements.
Sized against real sea states.
Our energy model runs on NOAA buoy records from station 42035, 22 nautical miles east of Galveston, 2023 to 2025. The Gulf is not the North Sea: typical waves are under a meter. The design has to earn its keep in ordinary conditions, not in storms.
- 0.82 mMean significant wave height
- 1.40 m90th percentile
- 5.45 mLargest recorded
- 33,668Observations
Show as tableHide table
| Month | Mean (m) | 90th pct (m) | Max (m) | Observations |
|---|---|---|---|---|
| Jan 2023 | no data (buoy outage) | |||
| Feb 2023 | no data (buoy outage) | |||
| Mar 2023 | 0.98 | 1.36 | 1.60 | 283 |
| Apr 2023 | 1.01 | 1.47 | 2.47 | 1,434 |
| May 2023 | 0.64 | 1.16 | 1.84 | 1,476 |
| Jun 2023 | 0.76 | 1.23 | 1.68 | 1,438 |
| Jul 2023 | 0.65 | 0.98 | 1.39 | 1,487 |
| Aug 2023 | 0.70 | 1.11 | 2.53 | 1,482 |
| Sep 2023 | 0.55 | 0.90 | 1.33 | 1,422 |
| Oct 2023 | 0.92 | 1.52 | 2.09 | 1,150 |
| Nov 2023 | no data (buoy outage) | |||
| Dec 2023 | no data (buoy outage) | |||
| Jan 2024 | no data (buoy outage) | |||
| Feb 2024 | no data (buoy outage) | |||
| Mar 2024 | no data (buoy outage) | |||
| Apr 2024 | no data (buoy outage) | |||
| May 2024 | 1.02 | 1.37 | 2.27 | 959 |
| Jun 2024 | 1.00 | 1.86 | 3.19 | 1,440 |
| Jul 2024 | 0.76 | 1.05 | 5.45 | 1,460 |
| Aug 2024 | 0.46 | 0.78 | 1.29 | 1,486 |
| Sep 2024 | 0.75 | 1.69 | 2.75 | 1,433 |
| Oct 2024 | 0.87 | 1.44 | 2.40 | 1,487 |
| Nov 2024 | 1.07 | 1.76 | 2.51 | 1,437 |
| Dec 2024 | 0.98 | 1.46 | 1.82 | 1,488 |
| Jan 2025 | 1.09 | 1.71 | 3.37 | 1,468 |
| Feb 2025 | 0.84 | 1.54 | 2.12 | 1,342 |
| Mar 2025 | 0.89 | 1.49 | 2.38 | 1,487 |
| Apr 2025 | 1.03 | 1.78 | 2.26 | 1,428 |
| May 2025 | 1.05 | 1.66 | 2.37 | 1,483 |
| Jun 2025 | 0.87 | 1.14 | 1.83 | 1,437 |
| Jul 2025 | 0.68 | 0.92 | 1.39 | 1,484 |
| Aug 2025 | 0.43 | 0.71 | 1.29 | 1,478 |
| Sep 2025 | 0.62 | 0.91 | 1.61 | 699 |
| Oct 2025 | no data (buoy outage) | |||
| Nov 2025 | no data (buoy outage) | |||
| Dec 2025 | no data (buoy outage) | |||
Early. Here is exactly where.
We would rather you know the stage than guess it: pre-revenue, pre-seed, student-founded at Texas A&M. This log is the whole story so far.
- DoneArchitecture chosenFloating ladder frame. Five-bay prototype module modeled in CAD.
- DoneEnergy modelBuilt on 2023 to 2025 NDBC buoy records off Galveston (Fig. 4). Outputs are screening estimates, not measurements.
- DoneGood Bull Pitch winnerTexas A&M McFerrin Center for Entrepreneurship, October 2026.
- NowLab conversationsEarly conversations with marine robotics groups at APL-UW and Oregon State.
- NextSingle-bay bench rigThen a five-bay tank test.
Three undergraduates at Texas A&M.
Siddharth Seshampally
Co-founder · Oceanography
Product, physics, and customer conversations.
Aarya Bookseller
Co-founder · Computer science
Software and controls. Minors in physics and statistics.
Salvador Magaynes
Co-founder · Mechatronics
Mechanical design and electronics.
Tell us what a cut-short mission costs you.
If you operate AUVs, run a marine robotics lab, or work on offshore infrastructure, we would like to hear how you recharge today and what it costs. A reply from you shapes what we build next.