Home        Store        Docs        Blog

Blue ROV to the next level

(Oystein Skarholm) #1

OK, I have been using the BR2 professionally for quite some time now and performed inspection of a sunken boat in harsh ocean conditions and I have inspected more that 7000 meters of pipeline from -1 to - 60 meters in open water conditions, drill rig anchor lines in fjord conditions, fish farm anchor lines, 7000 m^2 seabed survey, internal pipe inspection through a 225m long inlet pipe, inspection inside a hydra plant tunnel and more to it.

I would say that the little worrier BR2 really stands out like true piece of professional kit. BUT, I want more ! I think that this product is too good to just be a camera (great camera the new low light!!)
Guys you HAVE to take this to the next level, this is just too good to be a simple underwater observation drone… I have some thoughts of what I miss in order to make the best small ROV even better, and turn it into a tool package. It will off course be a “new” model based on the same proven technology.

Balance and GOG:
When operating at some depth and current the ROV tilts nose when going down vertically because of the tether drag. I the tether is mounted on top, the nose will pull up due to tether drag when going fwd. A way to reduce this is to make the ROV slight larger and heavier adding larger vertical stability and decrease the tilt momentum created by the tether. This will also allow for more smooth operation of the ROV.
Normally I do all the fine maneuvering on 25% gain, but if current and tether drag gets the best of it I have to go to 50%. That is fine for straight flying, but when stopping for closer inspection and trying to rotate etc, the ROV moves to quick, and at the same time, I cannot go down to 25% as the current will pull me off. In summary, a larger and heavier ROV (open to allow water flow through) would give more stability and smoother operation. Im thinking closer to 30Kgs then 20.

Control of video:
I know this is being worked on already but here is a big area for improvement for bringing the BR2 to the next level. As of now I use a screen recorder to record the video to get the OSD to show on the product. The screen recorder also inputs text and time/date on the video. The trouble is now that cannot do anything on the computer as this will show in the inspection video (Editing after is extra work, time and money)
The ability to receive signals from several cameras and split the signals to separate screens would be a great benefit and a great tool. The QGC should be a control page that acts as video matrix that sends the video links to the different ports. You would then choose from QGC which camera to record - including OSD.

Auxillary equipment:
The option to add extra lights, motors, manipulator/grabber, pump, or other instruments and control them from QGC would be awesome. As of now, when I use the acoustic positioning system, the transponder gets power from the ROV as soon as I turn it on. (not good for the transponder to be on in air)
Lets go with the idea that the ROV i built larger as suggested, adding a pan tilt, zoom? camera would mean a world of difference in some tasks. Imaging controlling this from a menu in QGC.
Or a grabber of 5F manip, powered on/off and controlled from QGC ( manip controlled via extra joystick but interfaced through QGC)

When going deep or going where there is current, or both at the same time, is a true challenge for small ROVs as the tether drag takes over. To have the tether reel sub sea and able to pay in and out would mean the world in difference. Imaging power and signal umbilical coming from the surface, down to the reel, The tether-reel being heavy enough the hang more or less straight straight down. This means a short tether to the ROV with less drag. The TMS existing in the marked today is very expensive and mostly exists for larger "professional" ROVs. But I see and read so many clever people in here that Im sure someone could come up with a brilliant idea for this. My current solution is a “heavy” shackle connected to the umbilical 25m back from the ROV to account for the tether drag, but I know what I would prefer :slight_smile:

Two function grabber or a simple manip:
is MUST have. I know there are several small brappers out there, but can they be easily hooked up to the BR2 for power and controls?

I know I am asking for the moon but miracles have happend before :slight_smile:


Hi Im Christian ..Pro ROV pilot and Diver.. 100 questions.. anyone up for it?;)
(Etienne Demers) #2

Hello Oystein,

Thank you for sharing your experience on this forum.

For the control of video, have you considered using IP cameras? I have mine going through a router and then through the Fathom-X board. It then goes to my osd/recording software and live stream to a secure website for clients…

Also you can send the BR2 camera to several places at the same time. I am presently discussing this in another topic. “Get video in Videolan”

Anyway, I just got the BR2 a couple of weeks ago and I am building a surface controller. With your feedback, I am now thinking that I will add a hardware gain control for the rotate signal since QGC does not yet permit this in its setting.


(Oystein Skarholm) #3

Thank you for the answer. Since I am an oldtimer, with probably to many hours under water, I am more into the actual performance of the work and unfortunately not very familiar with routers and IP / UDP etc. But what you are describing is interesting. However for someone like me to be able to do this I need a simple setup guide and description of what hardware and software to use. It would be good if BR actually made this as an package to buy from their store. For me it would be a box to connect to before connecting to the FX card.

(Kevin) #4

@SDI Good post on the requirements and a wish list for future affordable ROV technology. Not that long ago (2014) our only option was the small OpenROV kits with 100m depth rating and almost no tooling. Or you had to go build your own. Now, we can do 300m and lots of room for tooling and camera stability. A huge improvement over what we had.

If you wanted to scale up into a platform that is larger and more powerful, I think the limiting factor will be scaling up the propulsion units. The T200s are a good size for the BlueROV2, but if you wanted something larger and heavier, then I think an increase in propulsion power and larger thrusters would be required. And that leads into a heavier power system, frame, etc. At that point, it won’t really be man-portable and be difficult to launch/recover without the use of winches.

There might be more you can do with the existing frame that I will highlight below. I also have one of the older beta kits (I think #2 or 3 from the RPi variant) and I have been able to keep up with upgrades throughout the years. It is acting as a bit of a demonstration platform illustrating that you don’t need to buy a new ROV each time a new revision comes out, you can simply upgrade the one you have. Which I don’t think any other commercial ROV has the capability to do and one of the main selling points for me when I was looking for one years ago.

Balance and CoG:
Good observations on the balance and tilting issue. There are two upgrades I would recommend to make it more stable:

  1. Convert the standard 6 Thruster to an 8 Thuster variant by moving the vertical thrusters outside the frame and installing two more thrusters on the extra mounting holes. This configuration is already supported in ArduSub: https://www.ardusub.com/introduction/features.html I have driven this configuration and I really like it because you can trim the pitch and the roll and the autopilot compensates to keep you at your desired angle or you can remain straight and level.
  2. Add the payload skid and move the ballast trim weights from the ROV frame to the bottom of the payload skid. This will lower your CoG and should make it more stable at least for lower speed maneuvering. I don’t remember how it does at higher speeds with the skid. https://www.youtube.com/watch?v=gGHNgvnkq40

Control of Video:
How and where would you mount the extra cameras? With the main electronics tube pretty much full, any extra cameras will need to be mounted externally and pointing other directions. Which directions are you interested in looking at?

If more cameras are to be added, I think we’d start have to looking at bandwidth issues. You can only push so much video through the 80Mbps of the Fathom-X before having to add more boards, or possibly looking at a fiber optic tether connection. I don’t know of anyone who has a working fiber system that is affordable.

Auxiliary Equipment:
From looking at the Joystick Button Functions under relay Control and Servo Control, you have two relays and three servo channels to work with. https://www.ardusub.com/operators-manual/button-functions.html Using the “shift” option should open up more buttons for the Aux functions.

Operating a 5F manipulator is difficult just with buttons, so you would likely need a co-pilot with a dedicated “natural” feedback control system.

As far as manipulators are concerned, I’m following two options at the moment (found them on LinkedIn as some of the websites aren’t fully updated):
http://www.theblueprintlabs.com/ (I’ll take two of the Reach 4 Mini’s :slight_smile: )
I’d probably be looking to buy which ever one is least expensive if they ever put prices on their websites instead of me asking for quotes.

This one is most likely outside the scope of an inexpensive man-portable ROV. There is some heavy equipment required for TMS systems on the support vessel side, which I don’t think many of the BR clients (I could be wrong) have access to. I certainly have my own challenges just trying to build a light A-Frame for my side scan sonar. I know Nathan Perry for his deep video sled had some pretty powerful winch equipment.

(Oystein Skarholm) #5

Thank you for a good and thorough response. See below for answers to your question:

Scale up: Actually I find the ROV quite powerful and so far it has coped all work on 50% max. Scaling up the vehicle is just to make it more stable. With an open frame to allow water to flow through the frame I really don’t think it would be much of an issue when comes to hydrodynamics. Adding the payload skid will like you say, lower the CoG, but will it be enough although there is very little weighs added to the ROV. maybe a combo of 8 thrusters config. and payload skid will drastically change its flying capabilities. But looking at the layout of the 8 thruster version I would prefer the vertical thrusters to be further apart, again to increase momentum as this will demand less AMPS during operation. Putting anything ouside the fra is never a good idea (like the Lumen lights) as these are typical snag points for ropes, nets etc. and you will be stuck. I moved my Lumens inside the blue buoyancy covers.

I might try that :slight_smile: And even make the buoyancy blocks larger, - remove the plastic, and add more weight to it.
Video: The extra camera would be mounted outside the tube as that gives better visual when looking straight up and down. I would like to be able to see both back and upwards for the tether. The resolution of the video should be best at the channel you wish to record. The other channels are for information to help keep an overview during operation. An external pan tilt unit will enable you to check your surroundings without having to move the ROV. Its a good help fpr larger ROVs but I also miss it for the BR2 when operating in congested areas with lots of snag points. Job still needs to be done :slight_smile:

5F manip: Its normally operated by the ROV pilot - But even if it means to involve a co-pilot I think its a good idea. Don`t get me wrong, a 2F grabber will solve most capabilities of a small ROV 95% of the times.

The TMS in this scale would not be anything like what we have on ROV offshore vessels. it should be a simple reel with a subsea weight of say 25kgs that was capable of spooling in and out tether subsea. The topside cable does not need to be on a winch at all as this can be lowered or riced by hand or by using a simple lobster/crab pot electrical/petrol hauler. This is of course for use with a boat and gives noe meaning when operating from land.

Really liked the Reach 4 Mini’s…do you know the price (Should i be afraid to even ask?) Can any of these be operated through BR2 without to much fuzz?

Thank you.

(Jacob) #6

ACRO_YAW_P parameter will scale yaw input only :).

(Kevin) #7

I agree, a 2F manipulator is fine for most observation class ROV operations, especially if the ROV can hold different roll and pitch angles. A 2F works well in the water column, but I’d prefer to have a 5F because like to go plant on the bottom and then move the manipulator around for more precision. An ROV moving around trying to grab something is a bit like putting a grabber on a wrecking ball. Doable, but a little squirrly.

I’ve been thinking about how to address the control issue. The handheld gamepad controller is fine for working without a lot of AUX functions, but once you start adding a bunch of servo and relay functions, it becomes more difficult. You have to take your thumbs off the propulsion drive sticks to adjust functions. I feel as though moving to two joysticks with a bunch of buttons within reach of your thumbs would be better so you still have propulsion control.

I like the Logitech Extreme 3D Pro, but adding a second one for your left hand would be incredibly awkward. Maybe we can find two ambidextrous ones and get those integrated onto a plate or something.

I haven’t asked about the Reach 4 Mini yet, but I will. I’m not sure on the control either, it says RS-232. It will probably take some work to get it integrated into a BlueROV2 and QGC unless it works with PWM.

(Oystein Skarholm) #8


Better put on two, one grabber to hold the ROV and one 5F to work with :slight_smile: then we have micro WROV :slight_smile: I have asked for a quotation for the R4 Mini including the different configuration from 1 function, 2F through full range of motion.

Operating a F5 is ok to do alone as long as there is little or no current as I find the ROV very stable. And lets face it, if we want to more than grabbing on to something in mid water, the ROV is very light and will most likely move around. Sitting on the seabed makes it easy with input hold forcing the ROV down.

By the way, making the ROV slightly larger and heavier for stability reason also makes it easier to remove ballast when inserting payload skid with manip etc. I played with an idea (see below) how a longer version can be made with 7 or 8 thrusters. Also notice the extra flotation that allows for more weights for stability. Less water flowing through the ROV vertically will have little impact on the operation plus extra thruster(s) will compensate as well. The flotation is without fairing and epoxy coated only. it should be attached on 4 or 6 minipost with washer and R-clip for quick and easy removal and installation.

Sitting in the office watching these guys working with slightly larger ROV`s using F5 Rigmaster and Shilling T4…not to different from the R4 Mini :slight_smile:

(Oystein Skarholm) #9

BTW, do you have an external OSD/recording software? Can you share what that is?
Will the BR2 recording option be able to also include the OSD any soon?
Can you explain further on how you set up multiple cameras with some simple diagram ? Please ?

Thank you

(Etienne Demers) #10

Hello Oystein,

I can’t tell what BR will add to their QGC. For me, QGC is for the pilot. For survey and deliverables I am using IP cameras. I got the small payload skid from BR2 in order to mount additional hardware including cameras. I am also using their 3 inch enclosure for cameras for now.

Being that I am building prototypes, it keeps the costs down.

For the Software. Its my own. I developed it years ago with my first business and now I am tweaking it a bit.

I’ll post some screenshots when I am ready.


(Jacob) #11

If this does happen, it is not likely to happen any time soon. It will require significant development effort, and we have a lot of other things to focus on. In the meantime, there are many screen recording solutions available.

Features like multiple video feeds and text overlay are things that we are working toward right now.

(Oystein Skarholm) #12

Thank you for clarify. Can you explain a little how multiple video feeds gonna work, will the operator be able to see all at once or toggle between them.

Is there anyway to send the videosignal c/w OSD to another PC or monitor before it goes into the pilots PC. This way the recording can be done without catching any window switching the pilot may do between MAP and video for example.

I tried unsuccessfully to connect the Ethernet cable through a router to see if I could get the signals to QGC running on two PCs. Yhe idea was to have the map window open on one of them.

(Jacob) #13

This has not been sorted out yet, but I imagine that you may see each feed in a small window along the bottom, and click any of them to make them fullscreen.

The OSD is the QGC application, it’s not present in the video stream. You may route the video stream around however you want with a program called gstreamer, there is just a learning curve around how to do it.

What you may be interested in is running QGC on two separate computers to get the overlay and video for recording on one and piloting on the other. To do this, you will need to disable heartbeats from the ‘Mavlink’ settings on the recording computer, set an ip of on the recording computer, and change the gstreamer options on the camera settings page in the webui from

udpsink host= port=5600


multiudpsink clients=,

(Oystein Skarholm) #14

Thank you for this information, I believe I will need a switch or similar to split the signal from BR2 into both laptops?
Is there any you would recommend?

(Jacob) #15

This one has been my buddy at my desk for a while, but really any that you get should be able to do the job.

(Paul Phillips) #16

Hi Everybody,

Stumbled upon this thread.

@SDI, I am from The Blueprint Labs, we have been working with BlueRobotics and are very supportive of what they are doing. We have a BR2 ourselves. Next month we will be putting together a 5 function manip in the RS1 series. Currently we have a 2 function working on the BR2. There is a small python snippet that goes onto the PI that links it to our host software on the surface. It will be compatible with the QGroundControl in the near future but currently makes use of our own software (ReachControl).Single functions could also be made to take a PWM input if required.

To get a sense of scale here is the Reach4Mini on a BR2.

In the pipeline is an extra camera and light system with pan, tilt and zoom for performing close up inspection.

Please contact me directly at p.phillips@theblueprintlabs.com if you have any questions.


(undersearobotics.com) #17

@PaulP - I’ll ask the question I’m sure everyone wants to know. How much?

(Oystein Skarholm) #18

Thanks for answering here. I have tried to get in touch with you guys since for a few days now through your web page form and even by email. No luck so far, I will be sending you an email then :slight_smile:

(Roy Petter Dyrdahl Torgersen) #19

Hi Øystein, this is a modification we made for one of our clients:

This was designed specifically to have room for our EnviroSense system, which you can see part of on the starboard side and just above the battery tube; there is also an echosounder on the port side. We also designed it so that we can have room for up to 3 batteries, or 2 batteries and a manipulator, or whatever combination works best for each use case. We have made 3 of this design so far, and we are now working on improving it over Christmas, based on feedback received from our customers:

  • Adding another T200 truster aft for improved balancing / COG and lift capacity
  • Adding a transversal reinforcement forward to improve stiffness.
  • Changing the mounting angle of the vertical thrusters to improve flow, similar to that found on larger WROVs.
  • We’ve added a tether disconnect after this picture was taken.

What do you think?

BlueRov2 8 thrusters configuration
(Roy Petter Dyrdahl Torgersen) #20

I second that pricing request.