I am a commercial diver for the USDA Forest Service. We recently purchased a BlueROV2 to assist in underwater inspections and data collection for our bridges, reservoirs and appurtenances. We have a need to acquire bathymetry data in some instances and want to have that capability without investing in a BlueBoat. My thoughts are purchasing a roof rack and payload skid, to which we would attach/incorporate a Cerulean Surveyor240-16 MBES, a MarineSitu C3 Stereo Camera, and an ArduSimple AS-RTK2B-F9P GPS location starter kit, all integrated through an Ethernet Switch.
Bluerobotics insists that the Surveyor240 is not compatible with the ROV, but AI and forum research says it is with a little ingenuity. They also want to steer me toward WaterLinked DVL and a Sonoptix MBES…totally not what I’m after. I realize the GPS will not work underwater, but for bathymetry we would be operating at the surface.
I can’t be the only one who has had this idea. Looking for input and advice on how to best utilize our ROV to collect minor bathymetry data.
This idea does sometimes come up but I’ve yet to be convinced that the pros outweigh the cons.
Pro:
Only one vehicle to manage
Higher bandwidth via tether vs. Wi-Fi
Cons:
Efficiency- only a single battery, 3-4x the motor count, and a tether you have to drag around your survey area. If that’s target region is larger than your tether reach, then you need to live boat (inherently tricky) or launch from multiple locations.
Accuracy- the GPS in the Blue Boat sits on a proper backplane and is only covered with water occasionally in the roughest conditions. It will inherently get better lock than a GPS located at water level on an ROV. You can add a mast but this quickly becomes something to break off, and adds more drag and entanglement points.
The surveyor multi-beam is well suited to create bathymetric maps. If you’re investigating a bridge piling, it will show you where it is and may catch topography around it, but a sonar like the sonoptix echo gives you an actual picture, which moves like video from a camera.
If you’re set on using the ROV for sonar scans, the omniscan FS is a good option- affordable, and with the motion of the ROV a decent scan is completed- a static image unlike that imaging sonar output.
Thanks, I’ve read the guide prior to creating this thread.
One goal would be to record and analyze scour and streambed migration through performing streambed cross sections upstream, downstream and at the bridge from abutment to abutment. Probably an even cheaper way to accomplish this is using the Ping2 Sonar Altimeter and Echosounder instead of Surveyor240.
To gain lateral spatial context, like mapping and tracking undermining at the leading edge of a pier, our team liked the option of the MarineSitu C3 stereo camera over the forward scanning sonar, noting water clarity restrictions.
Right now I’m leaning toward using the C3-BR with the Ping2 echosounder and Curlean DVL-75. We can try to lock on gps from a surface device, but many of our structures are very remote and gps is often nonexistent.
The ping 2, as a single bean sonar, will give very course bathymetry vs. the Surveyor because of it’s 25 degree beam. You only get one measurement per ping, as opposed to ~16 from the surveyor. The area you take the average depth of increases with depth as a result!
The C3 camera comes with software to allow image capture of stereo pairs, however you’d be on your own when it comes to processing this into a 3D model. I’m not sure how i will you’d be able to capture scans as a result… f you have concerns about water quality, a sonar like the omniscan FS will provide better results than a camera would in turbid water!
The DVL only provides relative location- you tell the system where it is at the start of a dive. This is fine for things like position hold, but if you need accurately georeferenced data you’ll need an additional localization system…
GPS is worldwide! It may not be possible to get a lock with a poor view of the sky, but unlike cell phone signal coverage is just as good on remote areas as populous ones…
The tracker 650 is a DVL like the waterlinked, it is cheaper but the position estimate drifts a bit more. I was actually referring to their omnitrack solution, or waterlinked ugps. Both work to give the ROV absolute position reference at about 1hz, so when coupled with a DVL you get the best underwater position estimate possible. However, that’s a lot of $ for performance a GPS, especially one on a surface vessel, will early exceed.
GPS on ROV is the way to go, I was comparing it unfavorably to a surface vessel GPS. Sonar data is generally only as good as your knowledge of the sonars position when it was collected!
The payload skid is not necessary for any of that hardware!
An ethernet switch is required for the C3 camera, however the B&W stereo image pairs are only useful in the premium version ($3k+) of Metashape… I’m not sure if adding color from the third center color camera on that setup is possible. What are your goals for the stereo camera?
The Tracker DVL seems unecessary for your intended application. For best accuracy, to do a “cross-section” bathymetry of a river, you would drive the ROV along the surface perpendicular to the flow. The single beam sonar data could potentially be captured with the simple ping survey application, although this guide is written for the BlueBoat. Will the 25 degree beam from that sonar provide useful detail?
ReefMaster is definitely not required for single beam data - it is just a bunch of x,y,z coordinates that QGIS or other free software can handle well.
Setting up the RTK GPS in an ROV is going to be challenging - you’ll need to make sure the antennas are depth rated, and use jacket-pass through wetlink connectors to bring the cables in from wet to dry.
I think a single or dual omniscan FS on your ROV would be a better fit - you’d get DVL-like performance from having 2, and output data that is much more useful for understanding the bathymetry?
Well, I have been given a $10,000 budget and a goal of getting our ROV2 capable of performing bathymetric surveys and recording/mapping/detailing 3D underwater issues around bridges and impoundment structures. How would you do it, regardless of it being “ideal”?
As the clip from the product page shows, this hardware will both track ROV position, and move it across the sonar map that is created as the ROV pivots. If you need to geo-reference the sidescan data, you could either tie an existing feature into a lat/long coordinate, or add a GPS for getting a fix at the surface.
One thing to note - this sonar outputs an image, not XYZ coordinates. ReefMaster is a good fit for overlaying this collected data on the map, and exporting to google earth or other map program formats.
If you need depth measurements, running a Ping or the Cerulean Surveyor on the ROV, while the ROV is navigating at the surface with GPS lock is probably the best approach.
Generally, a BlueBoat with Ping single-beam is less than $10k and a better platform for future bathymetric surveys - I’ve been evaluating the Omniscan 3D for the Reef and been quite impressed with it thus far!
That looks great! I’d only caution that that C3 stereo camera requires expertise in image processing to generate models, no “turn-key” software is provided. I’d also mention that the 5 port ethernet switch will have all ports occupied with your configuration, so our upcoming RadCam upgrade for the BlueROV2 wouldn’t have a place to connect! It is a 4k IP camera that provides amazing image quality - if you’re able to scale the 3D photogrammetry models captured with it, you may see better results than the stereo camera image sensors can provide - although that stereo pair should make such scaling unnecessary, actually achieving that is tricky in practice!
From a compatibility standpoint, the rear endcap only has a limited number of holes! Assuming your ROV has the newer, 18 hole end-cap, a heavy upgrade leaves only 4 open ports (8 thrusters, Bar30, Tether, 2x power cables, vent-plug, lumen lights.) The Omniscan use up 2, and the Ping2 another - if you have a gripper you would not have an available port for the C3 camera, and if not, you’ll be at full capacity with it installed.
