In just over a week (June 6th - 9th), a bunch of people from OpenROV are going to attempt to dive a set of specially modified deep-capable ROVs to a 50m-long shipwreck at a depth of 150m below lake Tahoe. We’ll be using a deployment architecture that we’ve been perfecting over the years that involves a very small boat keeping station over the dive site while the rest of the people on the expedition run the mission from a remote location via long range broadband radio. Since the mission control site will have an internet connection, we’ll be able to live stream the entire dive over the internet.
We’ve outfitted a small inflatable boat with long range communication equipment, a USBL acoustic positioning system, and a low cost dynamic positioning system which we tested at Lake Tahoe this week.
The entire system can be packed up to fit in a minivan.
Our purpose for building this system was to demonstrate that many of the capabilities one might think would require a large research vessel can actually be achieved with off-the-shelf parts that are more portable and much less expensive.
Here’s a breakdown of the equipment we used:
We got this equipment on loan from Tritech so we could experiment with acoustic positioning while diving the shipwreck. Although these systems are quite expensive, we thought it would be good to get a sense for how easy it would be to integrate it with our setup, and having specific position data while diving on the shipwreck would also be very helpful information to gather for planning future ROV dives.
Since the 2-series ROVs have several auxiliary output wires, it was easy to power the transponder using the ROVs onboard batteries (we rigged up a 5v-12v boosting converter internally), and we mounted the transponder to a piece of polypropylene that we cut to size on the laser cutter. ROV performance was pretty manageable despite the added mass and drag, and we found that the ROV actually tracked straighter due to the horizontal center of drag being moved further aft. The Transducer head was mounted to an aluminium bar that could be fastened to brackets on the transom of the boat normally used to hold wheels. When under way, the Transducer bar was folded up, and when on location, a cotter pin could be removed to allow it to pivot downward into the water.
In general, the system seemed to work (we were ale to get pings from the ROV and a relative sense of its movement) however the locations reported by the system often seemed to be inconsistent. We’ll likely need more time to experiment with this system in order to figure out how to optimize its performance, but it’s good to know that the general concept works.
This is basically the same system we used two years ago when we attempted to dive on the SS Tahoe, but we’ve made a more compact enclosure for the electronics and have also created a mounting system that more solidly attaches everything to the boat.
Just like for the USBL transducer, we attached the system to the boat with a rectangular cross-section aluminium beam fastened with cotter pins to a wheel brackets mounted to the transom. A larger fiberglass tube that all the electronics are mounted to fits over the aluminium beam and is able to slide up and down along their axis. When underway, the fiberglass tube was slid all the way down the length of the aluminium beam, but once on station it could be lifted to a desired height and held in place with a push-button quick release pin going through one of several holes drilled along the length of the aluminium beam.
Inside the electronics box (mounted below the ethernet bridge) was a 74Wh battery, a POE injector (needed for powering the ethernet bridge), a wireless router which allowed people on the boat to connect to the network and see what mission control was seeing as well as get internet access), an OpenROV topside adapter, and a USB hub which was used for distributing power to several of the systems. Also (not pictured here) we added a BeagleBone Black computer and ethernet switch which allowed us to connect the USB webcam (mounted below the box) to the network so mission control could see what is happening on the boat.
This system worked very well but there are still some improvements we’d like to make. The main thing we’ll work on is organizing the system more cleanly. Fitting all those electronics in such a small package was not an easy feat, and consequently wiring harnesses inevitably tended to get pulled out inadvertently when messing around inside the box. We’ve been discussing a small rack-mount system that could be used to organize everything better and keep wires out of the way. Finally, we’ve also talked about mounting a GPS receiver inside the box that would connect to the BeagleBone Black (which is in turn attached to the network) so that people in Mission Control would be able to track the boat’s location in real time. In high wind and heavy seas, it’s nice for the people on the boat to have minimal things to do, so allowing mission control to track the boat removes some of the responsibility of the boat crew.
Station keeping has always been one of the most challenging parts of boat deployments. Even if there is a slight amount of wind or current, the boat very quickly starts to drift off station and pulls the ROV along with it. Large research vessels designed to support ROVs often have something known as a Dynamic Positioning System (or DP for short) which involves bow thrusters and a ton of sophisticated onboard hardware to keep them on station. After reading an iBoats forum post sent to us from Keven Klemens, we realized that a similarly capabile system could be implmeneted on a small boat for much less money.
The way this system works is by using a secondary actuator attached to the shaft of the trolling motor which electrically rotates its direction as the motor spins. A wireless remote control has a GPS receiver on it, and when you put it in anchor mode it controls the direction and thrust of the motor to keep you from drifting. Just like we saw on the original IBoats forum post by saabsaviorlee, we made a wooden platform for the motor to be mounted to that it could be strapped to the bow of the inflatable boat, and we brought a deep cycle battery with us that could be strapped below the platform to keep it from sliding around when underway.
I have to say, we were really impressed by this system. It worked straight out of the box, and when we told it to stay at a particular spot, we never saw our GPS coordinates drift by more than a few meters, even in wind and chop.
All in all, we’ve been pretty satisfied with the setup we’ve developed. Our hope is that by making posts like these, other people will be able to create similar systems which will allow more sophisticated dives to happen without sophisticated funding. There’s a lot to discover down there and the technology readily available these days can allow us to explore it.