Research
A laboratory
that sails.
Four research axes, carried out aboard HYDRO-01 and then L'Hydronaute: assisted navigation, retrofit, energy and water quality.
L'Hydronaute is open to researchers, laboratories and institutions. The data produced is versioned and destined for an open dataset; research missions are offered at a reduced rate.
Propose a missionResearch programme
The HYDRO-RD-01 programme runs four axes, aboard the V1 demonstrator and on a small-scale autonomous mini-demonstrator.
- 01
Assisted navigation & onboard intelligence
Experimental
An autonomous mini-demonstrator (ASV) is the test bench for assisted navigation: an ArduPilot/MAVLink stack tested in a basin, then on the Erdre. Human navigation data from Trashure Island feeds a behavioural-learning training corpus.
- 02
Reversible interface & retrofit method
Integration
An exoskeleton-to-hull connection that transmits propulsion loads with no permanent modification of the host boat — and a documented method, transferable from light launches to barges, with no heavy shipyard work or extended downtime.
- 03
Hydrodynamics & energy architecture
Integration
Three open questions: characterising and reducing the drag added by the floating exoskeleton, sizing the solar-battery-propulsion chain under variable load, validating the extrapolation toward larger boats.
- 04
Microplastics & water quality
R&D
Microplastic sampling and filtration carried aboard L'Hydronaute, in real conditions on the Erdre, with a protocol aligned on reference Loire-basin work — plus water-quality measurements (conductivity, temperature, turbidity) and biodiversity observations.
Propose a mission
Researchers, laboratories, PhD candidates, institutions: run your measurements or experiments in a river environment.
- Platform access at a reduced rate for research missions
- Onboard sensors available (energy, navigation, water quality)
- Collection protocol co-built with your team
- Data produced shared under an open licence
- Field: the Erdre, Natura 2000 zone, 15 km between Nantes and Sucé-sur-Erdre
Plate · Symbiosis 05 · Lichen — alga & fungus
A lichen is not a partnership: it is two organisms — an alga and a fungus — become a single body, able to live where neither would survive alone.
Our research works the same way: science, engineering and design fused into protocols no single discipline would produce on its own.
Citizen science
Trashure Island
A 3D navigation game that serves as our model: watching how players react and steer — movements very different from an autopilot's. Every game produces observation-action pairs, which we will compare with how our surface drone pilots.
Play Trashure Island Read the methodologyOpen dataset
Open baseline. Protected research. Our baseline data is meant to be open where it creates public value; partner teams' datasets remain protected; telemetry and participant data are kept separate and handled in accordance with the GDPR.
- Microplastics · Erdre — georeferenced samples and collection conditions. Publication planned late 2026
- Navigation · ASV logs — timestamped ArduPilot logs, Git-versioned. Publication planned 2027
- Trashure Island · corpus — observation-action pairs, documented schema. Publication planned 2027
- Protocol · EXO-01 validation — read the protocol. Published — v2.0, August 2026
State of the art — the barrier is operational
The European project HydroNet (FP7, 2008-2012) demonstrated surface robots for water monitoring, yet its own reporting concludes they “lacked robustness for extensive operational deployment by environmental agencies”. The barrier is operational, so we remove it by method — a validation protocol published before any campaign — and by operating the platform rather than delivering it.
Applications
- Microplastics · sampling, filtration, counting and characterisation in river conditions
- Water quality · conductivity, temperature, turbidity, dissolved oxygen
- Biodiversity · bank observation, acoustic monitoring, species presence
- Solar navigation · real-world range validation, energy profiles, use conditions
- River retrofit · modular exoskeleton transferable to other existing hulls
- Open data · versioned datasets, metadata, progressive publication
- Assisted observation · onboard vision and AI to annotate navigation sessions
- Education & citizen science · workshops, schools, local groups, Trashure Island game sessions
Team & collaborations
- David Aznaurov · machine learning researcher (M:3L, Institute of Mechanics, Armenia) — autonomy, point clouds, satellite calibration (Google Scholar).
- Elven Villecourt · founder — field operations, institutional relations; Head of Media, Paul Watson Foundation UK.
- Arpi Mangasaryan · architect; studied EXO-01 in her thesis and proposed an alternative version we use as advice.
- M:3L — Mathematical Modeling & Machine Learning Laboratory · partner laboratory (memorandum of understanding, 2026), Institute of Mechanics, National Academy of Sciences of Armenia — 3D, satellite data, applied AI.