Narrow passageway navigating ROV

[skyground]

Hello

Designing watercraft is quite a new experience for me and I am familiarizing myself with the theory of ship movement at the moment. I need to design a small and autonomus ROV (12 inches or 30 cm long), which needs to navigate in a narrow channel containing obstacles. It needs to do so as fast as possible. Some of the obstacles may be "unexpected" so the ability to accelerate, stop and turn rapidly on a moments notice is needed. To this end I so far have two ideas for the overall design.

1. Taking the hull of the S-100 class Schnellboot and adding a single high power waterjet (with reverse cup) at the stern and two weaker jets for steering at the bow. A narrow and smooth shape with the rapid response of waterjets might serve me well.

2. A catamaran. In the middle of the floats a 3 or 4 paddle (articulated?) wheels in series. Jets for steering (or a pair of waterwheels at the stern and bow, perpendicular to the ship movement axis). This would be easier to make.

3. A more or less conventional submarine with rudders for steering. The ability to dive would solve quite a few navigational issues. However, I have been unable to find an affordable underwater sensor solution. The nearby deadzone of the cheaper end of ultrasonic sensors is an issue. Machine vision camera and a projected light grid is an option I hope to avoid.

I have two questions for the audience here:
1. I would like to hear your ideas for the design of such a vehicle.
2. Recommendations for books or other materials regarding ship control theory.

 

[bucher]

Hi there. Sounds like a great project. I was involved in a couple of similar things in University.

Before I go into the specifics of the designs you've mentioned, I have a few comments on the controls for something like this. The paddle wheel design sounds cool, but pretty complicated, and has it's drawbacks (stopping/accelerating quickly for one). As for water jets, on fast full scale vessels they are great in terms of efficiency and manoeuvrability, but will be difficult to apply to such a small vessel. There may be some pre-made options, but consider carefully if they will actually provide you much advantage over a simpler propeller set up.

Propellers will give you all the speed you need and there are ways of making a propeller driven vessel very manoeuvrable. Simplicity and ease of building/fixing is often key to producing a quality final product in these sort of projects/competitions

The following are my thoughts on the designs you've proposed:
1. Arguably the easiest hull to build (one hull instead of two etc.). Although, that depends on the hullform a bit. Will probably be narrower than the catamaran design, which could help with navigating the course, however the cat design does come with its advantages.

2. Catamarans are good for this sort of project for a few reasons.

• They are typically fast and efficient hull designs
• They give you a large platform to carry the processor, batteries, sensors etc.
• The two hulls will make the vessel more stable than a single hull design. Therefore, the vessel won't roll from little waves or when taking corners as much. This is quite advantageous for a vessel navigating autonomously.
• Will travel in a straight line better than a single hull vessel (as long as the catamaran's demihulls are identical or close to). Once again this is good for an autonomous vessel.
• By having a propeller (or water jet) at the stern of each hull you will be able to almost spin on the spot. This works by putting one in forward and one in reverse. This combined with a couple of rudders should give you all the manoeuvrability you need. If you really need lateral movement there are ways of doing this with twin water jets at the stern instead of propellers (I imagine this would be tough to code up though) or you could add some kind of bow thruster. You can use the same control set up on the single hull design, although it might not quite be able to spin on the spot.

3. The advantage of this design is pretty clear. You can go underwater. This certainly could be great for avoiding obstacles. However, it complicates things quite a lot:

• You need to have at least some of the hull be completely watertight
• Your control scripts now have to deal with going up and down as well as port and starboard (don't underestimate how difficult this is).
• As you alluded to, it limits the sensors you can use, and probably adds one or two additional sensors (e.g. a pressure sensor for judging depth).
• Fitting everything you need inside a watertight 30cm long hull would be quite difficult.
• You need some way to make the vessel dive and surface (probably hydroplanes), which means additional controls etc.
• You would also likely have to weigh the vessel down a bit to the point where it is almost sinking in order to be able to make it dive.

I'll reiterate something I said earlier. Simplicity is important for this design. There are main two reasons for this. Firstly, you want to be able to build the vessel to a high quality, the more complicated the design, the harder this is to do. Secondly, you have to be able to write the control scripts for this, which can be difficult. A more simple controls set up will make this job a bit easier.

Hope this helps! Feel free to ask any more questions you have. I've had a little to do with challenges similar to what you are describing and have a couple of friends who work with Autonomous Underwater Vehicles. So I'll try to help if I can.

 

[portacruise]

If simplicity is important, what about a single prop at the end of a trailing shaft with the motor at the other end and that motor mounted centrally on a 180- 360 degree horizontal servo controlled pivot using #2 design?

PC

 

[Stumble]

There are off the shelf RC boats that can hit 75mph for <$400, how fast do you need this thing to be?

 

[messabout]

Visit your nearest model hobby shop. You can find a boat that is almost ready to go for less than $100 U.S.. The little boats are radio controlled, can maneuver very well, will go astoundingly fast, has re-chargeable battery, is fun to play with. I have scratch built quite a few models similar to the one you want. They are all propeller driven.

The boat need not be a complex design. A scow like model will work nicely but one with a pointy end may be better at avoiding hazards or deflecting itself from whatever barrier it may encounter.

I warn you that you may get hooked on RC boat modeling if you do this project. That can turn out to be a lifelong affliction whose only cure is to build more models, which is not an actual cure.

 

[bucher]

If simplicity is important, what about a single prop at the end of a trailing shaft with the motor at the other end and that motor mounted centrally on a 180- 360 degree horizontal servo controlled pivot using #2 design?

Also a good idea. An azimuth thruster type arrangement would give you good manoeuvrability.

Side note: Your speed is going to be limited by how good your course finding code is. Autonomous boats are rarely speed demons. So, perhaps don't go too crazy when picking motors etc.

 

[portacruise]

Also a good idea. An azimuth thruster type arrangement would give you good manoeuvrability.

Side note: Your speed is going to be limited by how good your course finding code is. Autonomous boats are rarely speed demons. So, perhaps don't go too crazy when picking motors etc.

I use this thruster principle when threading between boulders of various sizes going downstream on river rapids at speed (where rudders are useless). The power unit is light enough to be held in one hand if needed, so my arm movement is like the servo or it can be set down into a swiveling steering socket. The thrust is precisely controlled once I pick my course to keep in the deepest, best boulder free area. The boat is a 6 ft. one man inflatable pontoon with pointy ends.

I get the idea that raw speed capability is not that important since he mentions a sub design, he just wants to go as fast as the obstructions will allow?

PC