Pipe and box culvert inspection platform

George Sutton

I am thinking about building a robotic platform to visually inspect large diameter pipes (18 inch to 36 inch) and box culverts (several feet in width/height). Visual observation of primarily the bottom of these are desired to help determine whether a pipe bottom is corroded and shouId be sleeve lined, etc. I plan to use a wheeled platform, perhaps the Wild Thumper, with control by a Propeller. I will likely provide a tethered cable for use as a failsafe and also possibly for control and visual feedback. I would prefer to use some sort of wireless video system for short distance transmission, as the pipe or culvert will be perhaps 100 feet in length at most. I like the idea of a semi-autonomous robot platform, but it may be too cost prohibitive and too complex for my abilities. Any ideas or suggestions on what I may need to consider in the thinking phase of this project are more than welcome. Hardware suggestions will be most helpful to me. Thank you for any ideas.

SRLM

Adding a tether might be an issue, especially for such a small and light platform. Do you really need it? Why not do some tests with a transceiver and see what sort of reliability you get.

You could reduce cost by getting rid of the wireless camera, and replacing it with a video camera on the robot. Then, you drive it (by long sight or autonomous) for the distance of the pipe and return. You could then review the footage.

Ttailspin

This is how the 'big boys' get 'er done...http://www.inuktun.com/crawler-vehicles/
I know you were looking at wheeled chassis, but maybe you can find some ideas from that site.

Just a thought, Since you are using a tethered cable anyway, Maybe you don't need a 'Wireless' camera?

I have seen the operators of these things, pulling really hard on the tether cable, They pulled about 75' of cable into a straight line,
as they tried, successfully, to retrieve a stuck Pipe Inspection Robot..
I am sure you will be mindful of how you connect your tether to the back of your bot..

-Tommy

ElectricAye

Is visual inspection really necessary? Could you detect corrosion some other way - perhaps via electromagnetic inductive changes or sonic means? Transmitting visual information requires a lot of data transfer if it needs to be sent back to the operator and done in real time. On the other hand, if you could record data somehow and correlate it with your position, then you could operate autonomously. If you use inductive changes, your data recording/transmission requirements would be relatively small. Also, non-visual corrosion detection might be able to detect corrosion, etc. through sludge, etc. that would make visual inspection impossible.

George Sutton

Thanks, everyone, for the replies and comments. I am interested in exploring the bottoms of primarily corrugated metal pipe, which are frequently associated with collapses in fill embankments in the roadwork that I deal with. After so many years, the metal bottoms in particular rust away and this results in water seepage that leads to soil piping and sags or collapses. Seeing holes in the pipes would be very helpful in deciding whether or not to consider slip lining the metal pipe with HDPE or taking other corrective measures. Video transmitted wirelessly would be great, but as mentioned this would be costly. Perhaps on-board storage would be a consideration, to be viewed when the robot returns. I am guessing I could try using sonar sensors to look forward/sideways and upward to guide the robot along the pipe. Buildup in the pipe may be impossible to cross over, or the pipe may be collapsed at some point, requiring the robot to return. Tracks may be preferable to wheels, but I figure a wheeled platform would be cheaper. I am sure there are many other considerations I have not thought of. Further suggestions are welcomed. Thanks again.

NWCCTV

http://rctankcombat.com/articles/track-systems/ Good info on DIY tracks. I guess it all boils down to what your budget is. Here's one that would probably do everything you are asking (less the tracks) but it is pricey. They have several others. Here's the one with tracks. If nothing else you could maybe get some ideas on how to DIY.

Mickster

Not sure if Eddy Current would work in this application but this is an interesting thread:

http://forums.parallax.com/showthread.php/115551-Eddy-Cuurent-Analyzer-(Metal-tester)?highlight=Eddy+Current

Regards,

Mickster

George Sutton

SRLM wrote: »

Adding a tether might be an issue, especially for such a small and light platform. Do you really need it? Why not do some tests with a transceiver and see what sort of reliability you get.

You could reduce cost by getting rid of the wireless camera, and replacing it with a video camera on the robot. Then, you drive it (by long sight or autonomous) for the distance of the pipe and return. You could then review the footage.

I would like to hear more about any ideas you have regarding a transceiver. Any particular devices you may have in mind?

SRLM

There are a few transceivers that I'd like to try:
http://rfdesign.com.au/index.php/rfd900
http://www.digi.com/products/wireless-wired-embedded-solutions/zigbee-rf-modules/point-multipoint-rfmodules/xtend-module
https://www.digi.com/products/wireless-wired-embedded-solutions/zigbee-rf-modules/point-multipoint-rfmodules/xbee-pro-xsc

Note that these are all at 900MHz. You might also try some XBees at 2.4GHz to see if that makes any difference. You could also go the other way to 400MHz.

I have no idea what the radio effects of the pipe is. It may not do anything, it may bounce the signal around so much that it's useless, or somewhere in between. I'd buy a set and experiment on site. It may turn out that you don't need those powerful devices, it may turn out that one frequency is better than another, it may turn out that your antenna choice affects the signal, and so on.

If you're still interested in the wireless camera I've had good results with Team Black Sheep products: http://team-blacksheep.com/shop/cat:accessories

Gadgetman

What materials are the pipes and culverts made of?

A ring of ultrasonic transducers should be able to accurately test a concrete pipe for damage.

George Sutton

SRLM wrote: »

There are a few transceivers that I'd like to try:
http://rfdesign.com.au/index.php/rfd900
http://www.digi.com/products/wireless-wired-embedded-solutions/zigbee-rf-modules/point-multipoint-rfmodules/xtend-module
https://www.digi.com/products/wireless-wired-embedded-solutions/zigbee-rf-modules/point-multipoint-rfmodules/xbee-pro-xsc

Note that these are all at 900MHz. You might also try some XBees at 2.4GHz to see if that makes any difference. You could also go the other way to 400MHz.

I have no idea what the radio effects of the pipe is. It may not do anything, it may bounce the signal around so much that it's useless, or somewhere in between. I'd buy a set and experiment on site. It may turn out that you don't need those powerful devices, it may turn out that one frequency is better than another, it may turn out that your antenna choice affects the signal, and so on.

If you're still interested in the wireless camera I've had good results with Team Black Sheep products: http://team-blacksheep.com/shop/cat:accessories

Thanks for the input. XBee Pro with up to 28 miles LOS! And Team Black Sheep does have some good stuff, and an interesting web site. Thanks. I will have to research this more.

George Sutton

Gadgetman wrote: »

What materials are the pipes and culverts made of?

A ring of ultrasonic transducers should be able to accurately test a concrete pipe for damage.

Tusen takk! The pipes are metal, and the culverts are concrete. The ultrasonic transducer suggestion is something I have not considered. It may be very useful and I will look into it as a possible method. Thank you.