BUILD A SUBMARINE THAT DRIVES ITSELF
A resin-printed submarine that navigates underwater without GPS, using a Doppler velocity log and periodic surfacing for position fixes.
by CPSdrone
- difficulty
- ●●●●●
- time
- months
- license
- license not specified
- repo
- repo
●●●●● · months · license not specified · repo
WHAT YOU’LL NEED
Jump to section
COMPAREE VERDICT
This is a showcase of what is possible, not a project you can replicate without significant reverse-engineering. CPSdrone built an autonomous submarine that dead-reckons underwater using a 3,000-dollar Doppler velocity log, surfaces periodically for GPS corrections, and completed an untethered mission after fixing a magnetic declination bug that put the first run into the bushes. The red resin hull was printed on a Form 4L, the propeller is titanium, and a long-range radio lets it be steered home when it surfaces. The problem: there is no repository, no STL files, no BOM, no autopilot code, and no build guide. The YouTube video shows the result and explains the navigation approach, but you are on your own for every component choice, print setting, waterproofing method, autopilot configuration, and safety procedure. The single biggest trap is the sensor cost—this is not a budget build, and the key hardware that makes autonomous navigation work was bought, not built. If you have experience with autonomous surface vessels, resin printing for submersibles, and a reason to spend what a used car costs on a lake trial, this video will give you ideas. If you are looking for files to print this weekend, this is not it. The most likely failure mode is underestimating how much you will need to invent yourself.
IN THE REPO
GOOD TO KNOW
- —The Doppler velocity log sensor was purchased for approximately 3,000 dollars, not built.
- —Hull was printed on a Formlabs Form 4L resin printer; resin material and print settings are not documented.
- —Electronics, autopilot firmware, radio configuration, and navigation code are not published.
- —Testing was done in a shallow lake at approximately 1 meter depth; the stated Titanic goal at 3,800 meters is aspirational only.
- —No licence information exists; this is documentation of a custom build, not an open-source project.
Parts to buy
7 itemsFrom our check of the build. Exact quantities and part numbers are in the creator’s BOM.
Can I build this?
Build at your own risk. Projects involve tools, electronics and sometimes mains voltage — follow the creator’s safety notes.
Videos
Building a 100% Legal AUTONOMOUS Submarine (V3)
Shows the build, Doppler velocity log explanation, radio range test, first failed autonomous run (magnetic declination error), and second successful untethered mission. No step-by-step instructions or files provided.
More builds like this
All projectsGallery
Start here
Navigation into the creator’s own docs — we don’t rewrite the guide, we route you to the source.
- 1.Watch the main YouTube video to understand the navigation approach and see what worked and what failed (This is a showcase video, not a tutorial — expect to reverse-engineer everything.)
- 2.Source a Doppler velocity log sensor and verify compatibility with your autopilot platform(This is the single most expensive and critical component; budget approximately 3,000 dollars.)
- 3.Design waterproof housings for electronics, batteries, and the GPS mast mechanism(No CAD files or pressure testing data are available; this step requires submersible design experience.)
KNOWN ISSUES
- The Doppler velocity log sensor is approximately 3,000 dollars and was purchased, not built — if you cannot source or afford this sensor, the autonomous navigation will not work.
- No build files, STL models, or electronics schematics exist — you will reverse-engineer the entire project from video footage.
- The hull was printed on a Formlabs Form 4L, a large-format resin printer; smaller printers will require splitting the hull and planning watertight seams.
- The first autonomous run failed due to magnetic declination (true north vs magnetic north mismatch) — your autopilot must account for local declination or you will also end up in the bushes.
- Testing was done at approximately 1 meter depth in a lake; the stated Titanic depth goal is aspirational and not demonstrated — do not assume the hull or electronics are rated for pressure beyond shallow water.
- No published safety procedures exist for lithium battery flooding, loss of telemetry, or runaway navigation — you are responsible for every failure mode.
Are the STL files or CAD models available?
No. There is no repository and no files have been published. The hull design and all components are visible in the video only.
Can I build this without buying the Doppler velocity log?
No. GPS does not work underwater, and the DVL is what allows the submarine to measure velocity and dead-reckon its position between surfacing. Without it, autonomous navigation is not possible.
What autopilot firmware was used?
Not documented in the video. You will need to select and configure your own autopilot software and integrate the DVL, GPS, and radio telemetry.
How deep can this submarine go?
Testing was done at approximately 1 meter depth. The Titanic goal at 3,800 meters is stated as a long-term aspiration, not a demonstrated capability. Pressure rating is unknown.
Is there a BOM or parts list?
No. You will identify components from the video and source equivalents yourself.
Community builds
No community builds yet — be the first, we feature the best ones.
Discussion1
FROM THE COMPAREE TEAM
The Doppler velocity log cost approximately 3,000 dollars and is the sensor that makes autonomous underwater navigation work without GPS. If you were building this, would you buy the DVL or try a different navigation approach first?
CPSdrone
CPSdrone builds autonomous vehicles and has been iterating on this submarine design through multiple versions. The long-term stated goal is to one day photograph the Titanic wreck at 3,800 meters depth.
DISCLAIMER
- Comparee is not the author of the projects featured here. All rights to each project belong to its creator — every page links to the original source, and we never host creators’ files.
- Information is provided without warranty and may become outdated as projects evolve. Prices are indicative bands only — always check the creator’s parts list for current costs.
- Building and operating any project is at your own responsibility. Protective equipment, safe workshop practice and compliance with local regulations are the builder’s responsibility.
CompareeTEAM13d ago
The hull was printed on a Formlabs Form 4L, testing was done at roughly 1 meter depth in a lake, and the first autonomous run ended in the bushes because the map used true north while the compass read magnetic north. The Doppler velocity log sensor was purchased for about 3,000 dollars, not built, and is the single component that makes dead-reckoning navigation work underwater. The Titanic depth goal is stated as long-term aspiration, not a demonstrated capability. If you are evaluating whether to attempt this, know that you will be reverse-engineering every part from video footage.