User Tools

Site Tools


Remote-Controlled Submarine

Inspired by ,, ,

The idea will be to build a skeleton, then add a camera, then add propels and steering. It would also be really cool to add some form of sonar, this can probably extend the range of the underwater “vision”? A depth gauge is almost mandatory, then there should be a switch to make the robot automatically maintain the current depth (using the depth propeller).

Additional features, too cool to think about ATM:

  • Robot arm for grapping stuff
  • Sonar
  • Topology-sonar using webcam and line lasers. Could be used as an crude IMU for attitude determination/navigation, depth-measuring and for scanning the bottom-topology.
  • Metal detector/probing stick for determining what something is
  • Compressed air-container, allowing release of small amounts of bubbles to indicate position on the surface
  • Speaker, to play 8-bit music to the fishes (
  • Forward thursters ("bilge pumps") for fast forward!
  • Harpoon for fish hunting (Technically, the operator will need a fishing license. Food for thought: If we make the robot an autonomous hunter, will the robot need a fishing license?)


The more the merrier; please sign up if you want to help with the project.

Parts list

  • Plastic rør (Ø20-30mm ?) + samlinger
  • Undervandskamera, vandtæt til 20-30m (se, Vandtæt til 50m, inkl. lys, 3743,75,- )
  • Tryk-sikker kasse til elektronik
  • 1-200m CAT5 kabel
  • 1-200m strømkabel (2 ledere)
  • 12v batteri (bilbatteri?)
  • Propeller (1 dybdepropel, 2 styrepropeller)
  • Microcontroller (Arduino?)
  • Vandtæt samling til kabel attach/detach


The most common seem to be using bilges and converting to propellers.

Power budget for needed propellers (32 l/min models):

Part Usage (12V)
Up motor 2,5 A
Left motor 2,5 A
Right motor 2,5 A

Total: 7,5 A

Forward booster (62 l/min): 4 A

Blueprints for shaft converter :

We can buy the propels here:

Motor Control

It is suggested to use a H-bridge configuration with PWM (Pulse Width Modulation) to enable precise control of the thrusters. The H-bridge (L298) is driven by a PWM signal on its EN pins and the direction is controlled via its IN1 and IN2 pins. An Arduino is used as controller together with the small potentiometer-Joystick we have at the hackspace

3A H-bridge L298

Schottky Diodes

5V regulator e.g. 7805 Misc. resistors and capacitors


Pressure sensor/Depth Gauge

If target depth is 50m sensor needs to be able to go to ~600 kPa (~100 Psi)

Maybe SPD-100-G-2, e.g. from ELFA.

On Sea Perch they use Arduino as controller and the MXP4250 by Freescale, as presure sensor. We should be able to get two free samples from the webpage of Freescale. It is also available on, where then the price is at 130DKK.

UPDATE: We now have a MXP4250GP, thanks to Flemming!

Potential problems

Protecting the electronics

Both from water and from the pressure. One solution is to build a air-tight water-proof box, but then there is the pressure to worry about. Maybe it is easier to coat the PCB's in “lak”, as done by some overclockers when they watercool?

For creating watertight seal, see:

Interference on the cable

We will probably run both analog (camerafeed) and digital (steering, comm.) through the same, quite long, cable. Maybe we need to go to CAT6? Or two parallel cables?

Cat 5e STP 50m - 150,- Cat 5e FTP udendørs Netværkskabel 100 meter - 589,-


It quickly gets dark underwater, some form of preferably quite strong lighting is required.

Protecting the propellers

Stuff (cable, nastyness, etc.) in propeller = stuck robot. Maybe a form of “gitter” will do, but it might reduce the effect of the rotors as well.

Air in the skeleton pipes

The skeleton pipes should probably be airtight, to not increase the weight of the robot dramatically. This should also give the robot “updrift”.

Preventing rolling

The robot should be kept upright. The standard solution seems to be to have two “pølser” of light material (pvc pipe, plastic foam).

Randall Munroe (xkcd) seems to be doing without.

Work Sketches

Sketch of skeleton Considerations about frontal center of gravity


Initial design is to have an Arduino controlling the motors onboard, communicate with it over RS485 with a laptop.

Control Software

None yet.

Laptop Software

Initial GUI can be found at Leo at Launchpad. The software is naturally named “Leo”.

Pictures from the building process



Permalink projects/rc-submarine.txt · Last modified: 2013/04/08 15:47 by