First Underwater ROV Project


This is my first ROV Project which is already underway. Here is a list of specs.

  1. Dimensions: 42"L x 14"W x 10"H (excluding control surfaces and thrusters).
  2. Hull Design: Oval (similar to an old Russian Typhoon Class Submarine).
  3. Hull Material: Fiber glass & Carbon fiber.
  4. Power: Internal from 2 - 18AH agm batteries.
  5. Thrusters: 4- T100 Thrusters from Blue Robotics w/ESC.
  6. Control: Logitech Wingman 3D Pro Extreme.
  7. Lights: 2- 3000 Lumen Cree T6 LED’s.
  8. Camera: 1 PTZ IP Camera. Bow Mount.
  9. Sonar: 1- Raymarine Dragonfly 6 GPS & Transducer.
  10. Electronics: Arduino Mega, 2-L293D Motor Control Shields, 1-5 Port Ethernet Hub/Switch, 3 Servos(TBD).
  11. Ballast: Fiber glass & Carbon Fiber Air Tank, 4- Ballast tubes w/ rubber bladders and 2-12vdc Soleniods.
  12. Possible Addition: Digital Compass/Directional Heading.

I have begun the construction by cutting the basic hull mold out of stryofoam. I am in the process of coating the mold with plaster to further smooth and refine the shape. I have built the mold for the air tank and have all the fittings on hand. Next step is to fit the 2 thruster shafts, the rudder shaft, camera view port and light housings before fiber glassing and the carbon fiber finish. The camera view port is 1/2" thick acrylic plexiglass which I have to shape to fit the bow. I have already made a concrete cast of the bow to use for shaping the plexiglass. I did the preliminary design in CAD and have made all my patterns from that. As I am focused on inland fresh water lakes, I don’t intend to go below 100’ so I plan on a hull thickness of ~3/8". The air tank will hold approximately 100 psi and is also going to be ~3/8" thick. I am not sure how many ballast cycles I can get from a single tank, but I will determine that through testing. I will be cutting a ~12" x ~12" hatch in the top of the ROV. I will use 2" x 1/4" thick 6061 aluminum bar stock to make the hatch plate. Most of the items are on order or already in hand other than some of the more pricey items. I welcome any feed back or questions.


Wow! That’s really adventurous. I’m interested in the moulding, is your plug (pictured) going to be the mould, or are you going to take a female mould off it? Is the gel coat going to be inside or outside the finished chamber? If you’re moulding around the outside, will you gel after the laminate is done? I used to do a lot of GRP work and I’m very interested to know how it will turn out, it looks quite big so I’m guessing the top and bottom won’t withstand too much pressure. Good luck with it all.


Yes, this is the inside mould. I will be applying fiber glass to the outside then cutting the hatch. Once it is set, the mould will be taken out in pieces(destroyed). I have the original tin pattern if I were to build another. As far as strength, it should hold up quite well with the combination of fiber glass weave, fiber glass mat(alternating layers) and a final layer of carbon fiber. The gel coat will be on the outside, clear coat over carbon fiber weave. It is ~42 inches bow to stern. That will give me just enough room for the batteries, air tank, 4 ballast, ballast controls and the electronics and servos. It will be tight, but assembly is from the bow/stern towards mid-ship. As far as being adventurous, I truly enjoy the design challenge, as well as the construction. Here is a picture of the ballast air controls pre-assembled. I will tighten everything once I am to that point in order to adjust for spacing.

This has the 2-12vdc soleniods, a manifold, and check valve with various off the shelf fittings to connect it all together. Operation is straight forward. Solenoid #1 opens(NC) to let air in to the ballast bladders to ascend. Solenoid #2 opens(NC) to release air from the ballst bladders and reroutes the escaping air to the ballast chambers to exit the hull. The check valve simply blocks the water from flooding the rest of the air tubing. When I get ready for ater tight testing, I will adjust buoyancy at that point to have a slightly negative buoyancy with the ballast empty of air.


Excellent project, well done so far!
I wondered if you’ve used Tri-axial cloth, you can get some very strong ones.


I will definitely look in to that, thank you for the information. I have decided to use #90 Isopthalic Polyester Resin as it is considerably stronger than standard resin as well. We have had a break in the weather the last two days so I have been able to take the hull mold out side to work on it.


Very interesting. I know nothing about carbon fiber and fiber glass blends, but is there a concern with air pockets in the hull walls? From my limited knowledge, fiber glass is bad for deep sea due to air pockets, permeability, and so forth. What’s your target max depth? Nice design though.


You’ll need a good release agent, although as it’s an inside mould, not quite so critical. The most important thing will be to make sure the laminate isn’t resin rich - just enough to wet out, don’t be tempted to slosh loads of resin on to wet it quickly. Sorry if you’re already experienced with it.


Good Morning from THE Great Lakes State. My max depth should be under 120 foot as most of the inland lakes are shallow. As for air bubbles/pockets, I plan to use a roller to get the initial layers as smooth as possible. I have thought about hooking up my vacuum pump and enclosing the ROV in a large heavy duty plastic bag after have applied a thin layer of fiber/resin and some release film. We shall see as that step gets closer. I am sure that with sufficient care and diligence I could maintain water tightness to a depth of 200 feet or so. Thank you on the design as well. I have spent a great deal of time through trial and error in coming up with it. My first couple of designs were modular, but ended up being to lengthy to be of much use for maneuvering. They could obtain a greater operating depth though. In the end, I went with a design that would work given the parameters, such as inland fresh water shallow lakes, and maneuverability around obstacles, ie. tree trunks, sunken craft.


Exciting! reminds me of the WANDA AUV:


I had not seen this until you pointed it out. I like their approach as well. Having served 9 years in the USN, it is exciting to see the progress being made in AUV/ROV technologies. My last duty station was an AUW (Advanced Underwater Weapons) shop in the south pacific. I was able to learn a great deal about the technical difficulties of deep sea underwater design. But this was all before internet and cell phones (late 70’s early 80’s). I am truly impressed with both their design and the capabilities of their vehicle. Thank you for bringing it to my attention. I spent my first tour on an aircraft carrier and often thought about a fighter style craft design of a 1 or 2 man attack submersible. Just a young mans fantasy of designing and operating a vehicle that could “fly” underwater at high speeds without “crashing”. This present craft is purpose built to search and locate in shallow waters to aid divers in recovery. Maybe my next design will be a fanciful design from my youth.


@e4andy I have taken your advice and found triaxial carbon fiber in an 8-11oz weight which will do just fine. Thank you again for the advice. As for a release agent, I am still thinking this one through. Once the plaster is finished, I am thinking of applying a sealer coat (or two) and then a high gloss finish to aid in release. I am thinking that my first layer of two of fiber glass (or carbon fiber) will be hand rolled. Once cured, I will then look into trying a vacuum forming process for the final layers. This will give me a better weight to strength ratio when finished. If I can achieve a 60% to 70% resin content I would be very pleased. Final gel coating I plan on using an electric sprayer that I purchased. It is designed for thicker liquid applications( lower viscosity) such as deck stains. Once again, thank you.


Hi again. Can I suggest the areas that might cause issues will be getting the cloth to drape around the compound curves. The middle section won’t be a problem but the corners may keep lifting. I’ve never used carbon fibre but ordinary glass weaves can be quite challenging. With a good ribbed roller you should be able to get rid of every single air bubble if you work slowly enough. The trick is to use minimal catalyst so you won’t be panicked if it starts to go off too soon. It’s better to mix small batches of resin because a large volume in a bucket will start to set quicker. And when you mix a new batch, don’t use the old stirring stick, because the resin on it is part way into its cure. A little tip my first boss showed me, which works well if you’re using, say 16-20 oz paper cups for resin, don’t hold it like you would a straight beer glass, with your fingers wrapped round it, the warmth from your hand will start the cure. Hold it fingers under the edge of the base and thumb on the top rim, that way you transmit least heat. Believe me, the worst feeling is when you’ve got lots of expensive cloth cut & laid and the resin starts gelling!
Also, I would definitely recommend a proper release agent, just a high gloss won’t work, because the resin will bond to the surface at a molecular level… Even though it’s a one time mould you don’t want to spend hours chipping away bits of plaster from the inside. I would think you need at least 6 coats of a good release wax, Meguiar’s Mirror Glaze is great stuff. You could do 6 coats in a day.
Meguiar’s Mirror Glaze

Good luck.
p.s. One of the worst jobs I ever had was making GRP kitchen sinks for the British Navy’s minesweepers 40 years ago. They had to be plastic so we took a mould off a standard kitchen sink. It brings me out in a cold sweat remembering trying to get the cloth to lay into the little grooves!


@e4andy Thank you once again for valuable information.I have done a few items with fiber glass, but only a few. I would have not known about holding the cup. Thank you again. I have watched countless videos on moulding with fiber glass and carbon fiber. Your tips will be invaluable to the success of my efforts.It’s like building a house. Anyone can purchase a book on how to do so. But to build it to last and to make it look beautiful, you need the advice and experience of a professional. I know that the compound curves on the bow and stern will be difficult, so I thought of possibly using vacuum forming. I guess I will wait until I get to that point and see first how it goes. Luckily I am not dealing with small parts or ornate detail with fiber glassing the body. I figure if I can get a layer or two down and let it cure, then vacuum form the rest, I won’t have to worry about the styrofoam mould collapsing from vacuum pressure. The difficult part is the 6 stainless steel shaft tubes protruding from the form(3/8"). It will be stainless tube inside a stainless tube in order to keep the thruster shafts and rudder shaft aligned. I will use 1/4" steel shaft(disposable) instead of the inner stainless steel shaft. The stainless steel shaft will protrude around 3" beyond the body once it is formed. The rudder shaft can be solid, but the thruster shafts must be tubular to accommodate the thruster motor wiring. It will be challenging, but that is what makes it fun. Once again, thank you for the advice. Always a pleasure.


I’m sure it’ll work out fine. It seems like you have thought of pretty much everything. I keep thinking of little snippets. The final one is that you will need to put some wax additive in the final gel coat otherwise it will always be tacky. Gel only cures to a non sticky state away from air, i.e. up against a mould, so if you want to use it for a final finish, it needs an additive, most suppliers will have it.

Can’t wait to see the end product.


@e4andy I still appreciate the advice. I managed to align and drill the thruster and rudder shaft holes. Tolerance is within +/- 1/32". I set the shaft bearing/support brackets in place which will be fiber glass/carbon fiber-ed into place. The brackets will support the shafts, bearings and servos when finished. Alignment was critical to proper fit and operation when finished. The bearings, shaft collars and aluminum bar stock have arrived as well. I made a jig for centering the shafts on the hull, which will also double as a forming jig for the two pieces of bar stock to be used as the hatch plate. Two pieces of bar stock will remain unbent, two will be bent to match the curve of the hull. When assembled, they will be attached to the inside of the hull and the hatch will bolt down to them. Here are a few pics with the first coat of enamel sealer applied.

Side shot-


I have shaped the acrylic plexiglass to shape after cutting it. I also cut two groves along the edge for better bonding to the carbon fiber. After sizing the thruster mount, I decided that 1/4" diameter stainless steel tubing would not suffice for strength and have ordered 1/2" and 7/16" dia. stainless instead. The 1/2" dia. stainless steel tube has an inside diameter of 0.444 inches where as the outside diameter of the 7/16" is 0.4375 inches. This gives me a tolerance of 0.0065". That is something I can work with to get a good seal.


Update: The 1/2" precision ground stainless steel tube has arrived as well as the set screw collars. Also, the 7/16" set screw collars, and bearings as well as the Adafruit 16-channel 12-bit PWM/Servo Driver-I2C Interface have arrived. I have also ordered 4 - 3650 4370KV 4P Sensorless Brushless Motor with 45A Brushless ESC so I can program the Arduino Mega with L293D motor controllers and the Logitech Wingman Pro 3D Extreme Joystick. I cannot use the T100/200 Thrusters outside of water for programming without ruining the bearings(water cooled). So I bought the 4 brushless motors w/ ESC’s to do that. I also ordered 1-Robotis Dynamixel AX-12A servo. If this servo proves to be strong enough for thruster shaft control, I will order an additional 2-AX-12A servos. I am awaiting the arrival of the 7/16" stainless steel inner thruster shafts and a 7/16" solid shaft. The solid shaft will be in used to hold the 1/2" shaft in place during the carbon fiber hull construction thus keeping the 1/2" stainless steel shafts aligned. Once the carbon fiber is cured I can replace the solid shaft with the permanent 7/16" stainless steel thruster shaft tube. On a side note, I also had to order a soldering station as my current soldering iron is of no use when soldering pc board components. Hopefully the soldering station, 4-brushless motors and servos will all arrive about the same time so I can start the programming part of this project.