Written by Ken Wednesday, 20 August 2008 19:43

ROV Chronicles: Part 2

Design  philosophy

In the second part of our series on designing and building an ROV, we look at the design of the vehicle. The plan is to build an ROV that is intelligent and not limited by depth, and do it for a fraction of the cost of current commercial ROVs by using mainly off-the-shelf parts.

 

The key to achieving our goals is to choose components that can withstand pressure, so that there is no need to build an expensive, heavy and unreliable pressure-vessel. Since pressure increases by one atmosphere equivalent (one bar) for each 10 metres deeper the vehicle travels, a trip to “only” 100m depth (our target depth for the first prototype) would require the ROV to withstand 11 times the pressure it experiences at the surface.

 

At 100m, any fragile components with air gaps in them are likely to be crushed – this includes a lot of electronics. Luckily, there are normally alternative components – most modern solid-state electronics can withstand pressure since they are effectively a lump of matter with minor impurities.

 

By replacing any suspect components, we can do away with a thick-walled metal pressure-proof container, and instead use a thin walled box which will be filled with oil, and compensated to outside ambient pressure using a rubber diaphragm system. Unfortunately, its very hard to find out all the details of which components are pressure sensitive, so there will be a certain element of trial and error in the process!

 

The design overview

 

Although some details still have to be worked out, the basic design of our ROV is pretty clear. We are going to concentrate on the control, propulsion and sensor systems first, and once happy with their performance they will be integrated into a chassis.

 

Rather than build a dumb vehicle, we are going to include a computer into the ROV itself. Although the first prototype will use a tether for power and control, the possibility exists to, at a later date, make the ROV autonomous, or at least able to carry out some simple tasks such as navigation and station-keeping by itself.

 

Therefore, the “brain” of our ROV will be a small x86 compatible computer (probably a mini-ITX or nano-ITX motherboard). This will interface with the surface controller via standard Ethernet CAT5 cable, which gives our prototype a maximum depth of 100m (since that’s the maximum length of CAT5 you can sensibly transmit data down).

 

The on board computer will compress video taken from a camera and send it up to a laptop on the surface for the operator to view. The operator can send commands down to the ROV, and the computer will translate them into control signals and send them to the motors via USB interface cards.

 

The motors themselves will be brushless DC electric motors, probably sourced from the radio controlled model market. Being brushless, they can be left open to the water (with some minimal protection) which means there is no need for expensive shaft seals or pressure proof enclosures.

 

The motors will be controlled by electronic speed controllers, also sourced from radio controlled models – these take a pulse width modulation signal generated by the computer interface board, and convert it into the three-phase power the brushless motors use. Because both the motors and controllers are commonly available, the cost should be (relatively) low.

 

The electronics (in their oil-filled pod), the motors and the sensors will be mounted onto a chassis – basically a boxy frame designed to hold everything together. This will probably be fabricated out of marine plywood, cut to the right shapes, and coated with epoxy or polyester resin (although it might be necessary to do a fibreglass chassis if there are too many air bubbles in the plywood).

 

In order to keep the ROV the right way up, the top of the vehicle will be made buoyant with a special material called syntactic foam (basically millions of tiny hollow glass beads embedded in solid resin) which can withstand high pressures – although for cost reasons simple solid insulation foam might be used for initial tests. The bottom of the ROV will be weighted with removable metal ballast (probably lead) which will allow for buoyancy adjustment and trimming.

 

It all sounds very simple when written down like that, but there’s a lot of work to do matching up various components, testing, writing software and putting it all together. Especially given that we are hoping to spend less than £500 to get this ROV in the water!

 

Next time: Electronics, soldering and pressure-proofing

 

Discuss this article in our forums