Topic: ROV mounted side imaging sonar

[bigkahuna]

I've been thinking about installing a Humminbird side imaging sonar on my underwater ROV and I thought I might get some ideas and advice from the group here.  I know they install side scan sonars on working class ROV's for survey work but I haven't seen any micro ROV's with them mounted, instead it's typically forward scanning sonar.  My principal use would be for search and recovery.  I think that it might be worth while having, particularly in situations where I don't want (or can't) launch a boat.  My plan would be to mount the transducer on the underside of the ROV, thus giving it a clear view of the bottom.  I would also need to extend the Humminbird's cable to 250', the full length of the ROV's tether.  In operation, I would run the ROV in a circle above the target area and watch the scan for items of interest.  Then continue to spiral around the target in ever smaller circles while diving until the ROV is close enough for a visual ID.  I anticipate max depth would be 100 feet in salt, fresh or brackish water.  I'd guess most of the time we'd be in water no deeper than 30-50 foot.

Any thoughts?  I'd like to use the 798ci hd si combo as it is the smallest but will it have sufficient power to drive the transducer through 250' of wire?  Would I be able to use Dr. Depth without GPS data (the GPS would be ashore with the operator and not in the ROV in the water)?

[sonar2000]

BigK, rather than sonar for the rov you might consider just a radio controlled boat with sonar and RF back to the shore.
I forget who here was working on just that.
Chuck

[bigkahuna]

BigK, rather than sonar for the rov you might consider just a radio controlled boat with sonar and RF back to the shore.
I forget who here was working on just that.
Chuck

I thought about doing that, but I think keeping the sonar on the ROV would keep things simpler and faster / easier to deploy and operate.  Plus I wouldn't have to worry about surface effects, cavitation, etc.  Last but not least, I have a ROV that I could very easily retrofit a sonar on to so there would be nothing else to buy. 

How about signal drop issues over 250' of wire?  Would I loose a lot of benefit by not having the 800 khz capability?

[Roddy]

BigK, Done that, Been there. I have a Seamor insp class rov and 200' cable, color/black & white tv for/aft.

Added a HDSI 180 xducer with cat6 cable and Interphase twin Scope fwd looking.

Fwd looking finds the target, side scan maps it, manipulator grabs it.

No cross talk between the xducers and motors.

Fwd scan helps in low vis waters brown/black and search.

The Seamor is not a little Video-Ray type but a mid size insp class ROV.

System works well with one ROV control station (joy stick/monitor plus three (3) lap tops for CCTV and side scan/fwd looking sonar.

And a 1500 watt AC/DC gen.

Plus drop cams and tow fish.

One boat opr and one sys opr.

For Small lake/pond opens I tail gate it with 1000watt gen and a Land Rover at waters edge.

It can be real fun sometimes!

[bigkahuna]

Thanks Roddy, you're the first I've heard who is actually using sidescan on a mini ROV.  I'm familiar with Seamor btw althought I've not flown one.  Another possible glitch to my plan is that the HB sonar won't be able to use the GPS data nor the ROV's on board compass so will have to manually input heading/speed into the HB topside unit.  Plus I won't be able to use a mosaic software to piece the scans together.  How do you handle that with your Seamor?  For forward scan I'm looking at maybe a Tritech Micro DST, nice tiny unit but 12x the cost of a HB!  Any other tips on using the side scan during operations?  What types of ops do you folks do down in SC?

[Rüdiger]

Hi BigK

I think that it is not possible to use a mosaic software, the software shows only the motion from the gps-antenna and not the motion from the transducer.

mvh
Rüdiger

[Roddy]

Rudiger, what you say  is true, BUT GPS wave does pass Thur water some! 15'-30' not a unfounded. I have a GPS ANT mounted on two of my ROV's.

BigK, There is a lot of ROV nav and mapping FREEWARE on the internet try it. You can get a three axis magnetometer card on ebay for less than $100.00. This will give you heading, depth, pitch, roll, speed over ground And can used by mapping software without GPS. Cheep and EZ. Interphase unit $2500.00 US. w/o laptop.

Look into Interphase for FWD Scan sonar. I use their Twin Scan Hoz/Vet Black Box w/ a loptop and the loptop records everything. NICE!

[bigkahuna]

Rudiger, what you say  is true, BUT GPS wave does pass Thur water some! 15'-30' not a unfounded. I have a GPS ANT mounted on two of my ROV's  ...  You can get a three axis magnetometer card on ebay for less than $100.00.

I've actually got all those left over from an AUV project I never finished.  Problem would be getting the data up to the topside unit.

Look into Interphase for FWD Scan sonar. I use there Twin Scan Hoz/Vet Black Box w/ a loptop and the loptop records everything. NICE!

I looked at Interphase but I'm afraid those transducers would sink my poor little VideoRay.  I was looking at the 180se, so do you have the Ultrascan 90 or some older model?  So are you using the Interphase on your ROV's?  I was under the impression the transducer cable had to be pretty short, is that not the case?  If it's in the ROV, how many wires in the cable and how far can you go?  We should probably chat some time, sounds like you've got a lot of answers... 

[Roddy]

BigK, CAT6/7 work's.  Check your msg box, I just sent you one. Roddy

[oce_uri]

Hi all,

Like bigkahuna, I am also looking to mount the 798ci HD SI unit onto an autonomous vehicle, in my case a surface vessel to conduct bathymetric surveys and then ultimately construct a sonar mosaic.  My name is Ashley Hutchins and I am an Ocean Engineering student at the University of Rhode Island.  This is a part of my year long senior design project.  I am looking to find more information about the ethernet port on the back of the device.  How does it communicate?  UDP?  I am looking to connect the ethernet to our existing mission computer (currently a Netburner but may be transitioning to a Phidgets) or even another computer like a FitPC to store the data files that would be written to the SD card on the Humminbird.  We may even look to live stream this data to our remote shore station GUI which is currently how we alter the code of our mission computer and get returns of our heading, speed, etc.  Any thoughts on the data transfer rates this would require?  We currently have line of sight wireless communication of 3 nm and are working on implementing cellular communication as well that will hopefully give us increased bandwidth.  Also, is it all possible to integrate the Humminbird into the mission computer?  For example, can we control the functions of the Humminbird using code written on the Netburner?  Any advice or tidbits would be greatly appreciated.  Thanks. 

We also will ultimately use the same electrical systems when we compete in annual AUVSI AUV and ASV competitions.  Links are available to our previous years teams on our department webpage (www.oce.uri with links on the upper right).  A smaller model of the boat we are using is the one depicted under the Surface Robotics Team.  Again, thanks in advance for any and all help.

Ashley Hutchins 

[Rüdiger]

Hi Ashley

Your link dos not work, please try it again.

mvh
Rüdiger

[bigkahuna]

Hi Ashley,

Sounds like a very interesting project, would love to learn more.  Do you have a website set up yet?  The department's website address should be here I believe:  http://www.oce.uri.edu/

Have you taken a look at DrDepth yet?  Per (the developer) has managed to decipher the ethernet connection so he might be worth talking to if you haven't already.

[Roddy]

Ashley,

Sounds like you need a little one on one time with the tech's at HD.
As Paul (Bigkahuna) posted DrDepth works well.

Post some detail of your projet, and see if you can get some old heads working!

Good luck and keep us posted.

                                  Roddy

[oce_uri]

Wow, thanks all for the quick response.  Yes, I apologize I neglected to add the crucial ".edu" portion to the web address.  Our department webpage is www.oce.uri.edu as Paul stated. 

Roddy, does HD refer to Humminbird?  Every time I tried to contact Humminbird I kept getting stone-walled.  They said their tech people don't speak to customers and I finally emailed Greg Walters who frequently posts here on this forum and he just directed me back to here. 

Paul, my senior design does not have its own website, that website link is merely for the robotics team that will kick back up in the spring and summer using the modifications we make to the existing electronics system.  (For example, we may switch over to a Phidgets board rather than the Netburner we use now and a 30m LIDAR we are hoping to use to better our obstacle avoidance and object detection which is currently not the greatest ... team last year had trouble with a person in the audience wearing an orange shirt when they were supposed to be identifying red buoys.) 

I am really interested to hear about this Dr. Depth software.  I talked to Robert Gecy yesterday and he also recommended it to me.  If he has incorporated the ethernet directly rather than uploading his files from the SD card he would be a great.  I'll be sure to email them and try and talk to someone more about the ethernet. 

As for more details about the project, our boat is currently in shambles right now.  We ordered a new 10 HP water-cooled, two stroke engine that will be powered with a gas/diesel and electric hybrid power system.  We are using 2 12V car batteries (I believe) and another separate power system in with our electronics box that currently uses Li-ion batteries.  We have two trolling motors on the back with 30 lb thrust each.  The general model of the boat is depicted in the surface vehicle webpage (only difference is the size of the hobie cat pontoons).  Our boat has 6' pontoons but I believe their's has 4'.  Our mechanical tasks include:  developing a pulley system to connect the engine to the alternator, designing a generator system to charge the batteries, and designing a cooling system for the engine (we need to dissipate about 5 kW of heat).  Our software team is currently trying to learn the existing Netburner mission code so that we can by system operational and then incorporate a collision avoidance leg to abide by COLREGs when out on Narragansett Bay.  Vision team is trying to develop the LIDAR and couple it to a pan/tilt camera to generate a 3D world to then populate objects when we see and identify them.  This is a lofty goal we're not sure will be accomplished .... LIDAR's our best bet for that one I think. 

And ultimately, our main goal is to sonar survey with the vehicle.  We would like to send out the ASV to a known location we give with say four GPS coordinates corresponding to the entire area we would like to survey.  (Could be too complex, maybe just do a starting and ending point and conduct one line and see what our returns are.)  Ideally, we would like to then get a "live" return of the data to our remote shore station.  From the data stored on the ethernet, we will then construct a sonar mosaic of the grid/line we surveyed. 

Again, any thoughts, comments, tidbits of advice are greatly appreciated and definitely desired.  Thanks!

Ashley

[bigkahuna]

Sounds like a very fun project.  I don't know if you've seen this already, but Harald Molle posted a number of blog pages chronicling his autonomous survey craft project:  http://diydrones.com/profile/HaraldMolle  Granted the boat is many times smaller than yours, you may find some interesting bits of info there.

What type of comms are you using to communicate with shore?  Will it be a continuous stream of raw data or will the boat process the data and then send it to shore?

I've thought about building something along the lines of Harald's project, only using a semi-submerged vessel (imagine half a SWATH).  Never took it any further than just thinking about it though...

[oce_uri]

Thanks for the link, I'll be sure to check it out.  However, at first glance it, the website appears to focus on aerial planes.  What is Ardupilot exactly?  I've never heard of it.  For our comms, we currently have a  2.4GHz wireless hub and Wifi antenna with line of sight communication of about 3 nm.  We are looking into cellular communication now to alleviate the line of sight problem and to potentially get increased bandwidth to enable live streaming.  Currently, our remote shore station is able to receive live streaming of heading, speed over ground, lat, long, and a few other things ... I'll have to check what else but essentially they're just the data outputs from our compass and other health sensors about the vehicle.  Previous boats have not had the need to send data to shore or to process it, they merely had to process GPS inputs and some information from the vision system from which the operator could then make adjustments to the code, recompile it, and then send it back to the vehicle to alter its course/objectives.  For example, one component of the competition involved detecting a heated object and squirting water on it.  This "water cannon" was controlled on the remote shore station and could then be turned on when needed. 

For our potential cellular comms, we are looking into buying a Verizon wireless hotspot, an omnidirectional antenna, and then a dual band (800/1900 MHz) signal amplifier to boost the signal and essentially make our own Wifi network for the boat.  Thoughts on this?  We're relying on our proximity to shore in the bay to help with amplifying the cell signal.  I think Verizon and other phone companies are also looking to develop a cell network extending up to 200 miles offshore.  As to when this will be developed, however, is another story.  Our advisor would also like us to incorporate satellite comm for potential use should we travel far offshore (not on this boat but potentially for future generations).  For this we have an Iridium receiver, but the bit rate is so slow that we can't use this to send any data or even compile the Netburner code.  We plan on using it for an all else fails, send the boat home now.  Depending on the cell network, we might not be able to send live data.  Right now getting video from the camera is iffy and far too slow to have a continual stream.  I think we get a picture about every 3 or 5 seconds, but I could be wrong about that.  I know for sure we do not get live video from the camera right now though. 

SWATH idea sounded interesting, not sure how you have half a SWATH though .... We tossed around the idea of a SWATH at first for this boat because of the increased stability and decreased drag by having the pontoons submerged.  However, we decided against it because any modifications we would make to the boat would change the center of buoyancy and center of gravity, thereby affecting the SWATH design.

[bigkahuna]

Here's a link to a list of Harald's blog posts:  http://diydrones.com/profiles/blog/list?user=3um9qjnkq5rxo

You're right, DIYDrone's main focus is aircraft UAV's but a lot of good info (and tools) that can be used in AUV design / fabrication can be found there.  Harald built several autonomous surface craft based on the Ardupilot board, which is an autopilot / IMU / multiple other use board.  Harald wrote some code (that is published in his blog) that he used to drive is boat with to survey a small lake autonomously.  Although a much smaller scale than your project the goals are nearly identical.  He used a Humminbird SI sonar by the way.

My half SWATH concept would be something between a self propelled buoy and Woods Hole's Nereus AUV/ROV.  The idea being that the limitation of small surface craft is their motion in any seaway which limits their ability to reliably scan the bottom with hull mounted sonars.  By moving the majority of the craft's ballast (and surface area) to a submerged hull and only having a smaller portion at the surface you minimize that motion.  Another way of describing it might be a shallow swimming AUV.  Keeping it shallow allows it to keep in constant contact with shore.

By the way, I was on the design team for the first commercially built SWATH many years ago.  One thing we learned was that there is a minimum size limitation for SWATHs to be of any benefit in a seaway.  I also worked on the University of Hawaii research ship Kilo Moana, which was a SWATH and even on a vessel of that size, it was critically effected by even small changes or shifts in ballast.

[oce_uri]

Thanks for the link to his blog posts.  I'll be sure to read through those.  Like I said before it was just at a quick glance that I noticed the website primarily focused on UAVs.  As for the problem with small surface crafts and their limitations to reliably scan the bottom, I had thought about having sensors on there (IMUs) to measure the six degrees of freedom in the boat system and try and correct for that.  Therefore, whereas the sidescan would essentially be skewed based on the movement of the boat (pitch, roll, yaw, etc.) if we knew the value of each of those six degrees of freedom from when the sonar first released its ping to when it received it, we would then be able to "correct" for the skew and restore the distorted image.  From talking to other people, however, they seem to think this will not have a significant effect on the sonar returns and that it would be very difficult to correct for each and every return.

 As for your half SWATH, what exactly do you mean by it being in shallow water allows it to keep in constant contact with shore?  I personally like the idea of SWATH boats, especially in rough waters where wave action is significant, but I feel like their design should be for boats whose design are relatively constant.  I've never actually seen one in action though... Any thoughts on our comms systems?

[bigkahuna]

There isn't a lot of info out on autonomous boats so I've had to look in other directions for ideas.  That's probably why Harald posted his blog on the DIYDrone site.  He's written a book, btw, look at his 17th post for details.  Take your "self leveling" sonar transducer idea, you might be able to adapt a self leveling camera gimbal design to work.  I've seen several in the RC FPV forums.  Speaking of autonomous boat info, have you looked at the MIT paper published on their "SCOUT" project?  That paper was written 6 years ago but might have some good info if you haven't already read it.

I think for maximum sonar transducer performance a number of factors need to be considered:

transducer angle changes caused by vessel pitch and roll
vessel steering as straight a survey track as possible
transducer depth changes caused by vessel heaving
transducer height above the bottom
eliminate cavitation noise from props or the hull interacting with the surface

One idea would be to have your autonomous boat pull the transducer in a tow fish.  The problem with that, I think, would be if the towfish happened to snag on something you might loose the boat and everything else.

Another idea to eliminate all of the above influences would be to design a fully submerged AUV that could operate at a shallow depth (3 - 10 feet below the surface) yet had an antenna(s) that reached the surface for real time comms with shore.  The goal here would be to design the vehicle to operate deep enough so that surface waves would have minimal effect on it, yet shallow enough that an antenna could still be on the surface.  Even an umbilical to a buoy with antennas would work.  The challenge, I think, would be to keep the vessel at a steady depth.  Pressure based depth sensors, in my experience, aren't really that accurate but I haven't tested this idea yet.

My last idea, which really isn't like a SWATH at all except that it has two hulls, one above the water's surface, and another below.  This would be similar to the last idea except the submerged AUV hull would be permanently attached to the bottom of a surface operating hull with one or more struts.  Maybe a better way to imagine this would be a sailboat hull with a very deep keel, and at the end of that keel is a very large, torpedo shaped bulb.  I've yet to build a prototype so I'm not sure how well it would work, but in theory it sounds good. 

As for your comms plan, your current plan is how I was going to do it.  I had no idea you could get 3 nm range though, I was shooting more for 1/2 nm or less.  Your future plan to use cell phones for comms sounds like a good idea also.  There have been lots of DIY UAV projects using iPhones and the like that might be worth a look.

[Roddy]

Ashley and Paul, Have you checked on LOFAR for as a comm link? You can push all kinds of data and get 10-15 miles with low power.

[bigkahuna]

Roddy's post reminded me of something else I wanted to mention.  If you look at what RC boater are using for control comms it's VHF (27 Mhz to be exact) as they find that UHF can have issues due to its close proximity to water.  If you have an Amateur Radio license and your project is "non-commercial" you can try even lower frequencies, even some that might penetrate water to a limited degree.  The issue there, I think, might be with the bandwidth you can send at those frequencies.  But again, I'm just spouting ideas off the top of my head and haven't really studied any of these ideas thoroughly.

[sonar2000]

Before we get too far into this frequency thingy, remember that frequencies are allocated by the FCC so before using any frequency check for the proper allocation..dont use one that is not specified for what you are doing...
Chuck

[oce_uri]

Hi all,

Sorry for such a delay.  I had a death in the family and haven't been focused on this project in a while. 

Roddy, I am open to any form of communication, can you give me more information on LOFAR?  Is it a network of stations you use similar in concept to cell towers? 

Paul, I agree with all the problems you listed but unfortunately we can't really change our transducer location.  Our boat is a surface vessel and I think a towfish would be too complicated to control for the scope of our project.  I am, however, interested in your deep keel design.  We were looking to implement a keel as a part of a cooling system but nixed the idea for the same reasoning as the towfish.  This boat will be in shallow water conditions and we felt a deep keel would interfere with the operation of the boat itself.  Due to some logistical issues we had to send the sonar back so I haven't been able to test it's data, but what do you think about correcting the returns based on the heave and pitch?

Ashley

[bigkahuna]

Turns out someone else already built a prototype of my idea:

http://www.cctechnol.com/uploads/SURVEYSURVEILLANCE.pdf

[Roddy]

BigK, Whats the price of the 9500 model? Could use one lite side scan/rov work.$$$$$$$$$$$ too many toys in the yard! Roddy

[Roddy]

Ashley, Check out LOFAR on the web lot of info. Roddy