Help - Camera In A Smokey Environment?

One of our client requires us to propose a surveillance solution to monitor certain machinery on the shop floor. The problem is the environment is pretty smoky. The main objective is to read the machines display board to make sure its optimal operation (shop floor is mostly unmanned). Has any one installed cameras in such environments before? Any recommendation?


Tough application. Getting the camera as close to the display as possible seems requisite.

Is it possible to hang fans or configure ventilation equipment to keep 'line of sight' clear between camera & panel?

Is there some form of serial output on the panel/machine that you can pick readings from?

The tricky part with a fan is that if the whole area is smoky, while it's blowing smoke away from the camera, it's also pulling in more from behind it. You'd need to pull clear air from somewhere and blow it in between the camera and panel - almost the reverse of a bathroom exhaust fan.

Other than that, I'd agree, getting the camera as close as possible would be paramount. Also, a color filter to match the display color might help - for instance, if the display uses red LEDs, then a red filter would block other light (smoke generally being white/grey, it would tend to "carry" any ambient light). You see this idea with laser levels that come with red glasses for clearer vieiwing in bright rooms.

Ultimately, Brian is right, the best idea would be to see if there's some way to pull the information directly from the panel... maybe even see if the manufacturer has an option for remote display.

Brian/Matt,

My initial reaction was also to include a few exhaust fans. However, it would be of little help as the whole shop floor is smoky. I was thinking of something that can filter out the smoke.

The display consists of Red, Yellow and Green (graphs and figures). The client mentioned that the machinery manufacturer offers end-to-end solution that includes, camera and remote monitors, but that solution would be extremely expensive. Also, they would like to monitor displays from 3-4 machines across the shop floor, we were thinking of single PTZ camera with a good optical zoom. But this would move the camera away from the individual displays, which would mean a thicker blanket of smoke

Pushkar: thus why my suggestion was to pull fresh (clear) air in through a pipe or duct, and inject that in front of the panel or between the panel and camera. Admittedly, that's probably an extreme solution and wouldn't be 100% effective, but... something to think about.

I don't think the PTZ idea would work, because besides what you mention, the PTZ would need to re-focus every time it changes positions, something that will be tricky if looking through smoke.

This sounds like an instance where one has to balance expense vs. effectiveness: the simple fact is, you ARE looking at probably needing one camera per panel, mounted fairly close, with no guarantee of how well it will work depending on the air conditions. It could add up to quite a bit of expense once you include the number of cameras, filters, software... testing various different solutions will add up in materials and time as well. The manufacturer's remote-monitoring solution may be spendy, but it's something that's guaranteed to work (theoretically), because that's what it's designed to do.

Hi! Just remembered a company that specializes in enhancing picture quality.

See Lyyn smoke enhancement application.

I haven’t tried it, but it might be helpful ? I think they only work with analog but I’m not sure.

What I have seen by testing different IP cameras with built in functions for enhancing image quality, is that they actually can filter out for example some snowfall from the “true” situation. So the picture contains less snow then there really is. Not sure if it would work the same for smoke. The image enhancing functions looks at snow as “noise” in the picture and tries to reduce it, so I guess it’s a lucky side effect of the intended function.

Might be worth a try to how these type of cameras work for your specific situation.

//Snowman

Undisclosed, Lyyn works with IP cameras now. It's called their Griffin encoder and it costs $5,500 USD MSRP. See our Lyyn Griffin review. We have not tested it and it may work but it's going to cost and it's SD only (I believe).

Hi, we are strong in this area in Sweden!
Besides LYYN there is also another clever video analyst company here - Imint
And I think their solution is less expensive than LYYNs

See IMINT Vidhance

Jan, thanks, that's interesting. I had never heard of them before. I just sent them an email asking for more information.

You are welcome John
They are both small and relatively young companies with just 3-4 employees.
LYYN has a turnover of approximately U.S.D 300 000 and Imint little less. None of them have made any profit to date, they living from venture capital. But never the less - the products are interesting and I hope and believe that they have a future. The primary difference between the companies is that LYYN has developed hardware (algorithms installed on circuit boards), Imint developed software for managing IP video streams.

Matt, I was thinking to setup one PTZ with fixed guard tour to watch the displays. We can request the client to stop the machines for sometime (as the facility runs 24/7) and flush the smoke for the demo setup. If this doesn't work, then we may have to propose single camera per display. Also, just for my reference, what kind of filter you had in mind?

Undisclosed, the LYNN solution seems to be interesting. However, the cost might be prohibitive in this case.

Jan, thanks for the information. We would prefer a software solution like Imint. Their website also lists capabilities of object tracking which we require for a different client. I'll try to contact them ASAP.

You are welcome Pushkar
If you think I can be of any further assistant, the fact they’re Swedish (my neighborhood), just say the word then I see what I can do.

Pushkar, the problem with a tour setting is, the camera will still try to re-focus every time it changes presets - if the smoke is too thick, it may not be able to do that. Now if you don't need full 24/7 views of the boards but it's more just a matter of checking them now and then, then a missed focus now and then may not be a show-stopper, as it may have better success the next time the tour comes around.

Still, the farther the camera is, the more smoke you have between it and the target, regardless of how much you zoom it in.

As far as a filter, I was thinking a color filter that matched the color of the display lights - ie. red displays would get a red filter, so other colors are blocked (smoke generally being white or grey, it would contain a pretty even level of all colors across the spectrum). That said, you mentioned that the display use multiple colors, so a filter may not work anyway.

I heard back from IMINT (the Vidhance offering). There's no commercial off the shelf offering for using it with IP cameras / VMS. I am going to talk more with them but if you need something immediately, might not be a fit.

What I see no mention of is the smoke getting deposited on the lens/bubble/view window. Whatever it has, the camera housing will likely have to be cleaned often.

Enclosed is a list of websites to explain the differences in these cameras & why cctv cameras dont work when looking at fire smoke , unless specifically designed for that applications .

Residual Elements & gases in the air from the smoke particles .

You must look at it as a UV , Thermal Image , not a visual image

using spectual analsys to see the picture , not the visual

http://www.signifire.com/ipcam.html

http://www.flir.com/cs/apac/en/view/?id=42586

http://www.infratec.de/en/thermography/thermographic-automation/early-fire-detection.html

http://www.directindustry.com/prod/jenoptik-i-defense-civil-systems/infrared-cameras-for-spectral-thermography-65823-960677.html

http://www.directindustry.com/prod/testo/thermal-imagers-5240-355429.html

http://www.directindustry.com/industrial-manufacturer/camera-61080.html

Fluke , Fike , Xtrallis , Many More

Hello, Christopher.

The issue is not detecting smoke. He wants to see through a smoky environment to read a machine display panel.

The products you mention are fire/smoke detection cameras - not the aim of the OP.

The Problem is you cannot unless you change the display to a different output , immage visable thru smoke. Best Solution is capture the display , Feed to a Parallel output , and send to a remote display or computer.

Best to use computer to display conditions and respond accordingly and capture images sent to a computer for a alarm or status pt.

Scada Systems use this same concept. Build in Displays that allow the operator interactive response & review.

If you follow the discussion, Christopher, you'll see the OP has already mentioned that the equipment manufacturer has a remote-monitoring solution, but it would appear it's prohibitively expensive.

Matt, thanks... I didn't think of the refocusing problem. My understanding was that once the camera was setup (in a smoke free environment) it would come back to the same configuration regardless of environment being smoky. What a bummer!!!

Carl, that's a good point. We were thinking of using an external enclosure for the PTZ which would be at least IP66 rated (Dust tight, not sure if it is smoke tight). Of course the client understands that regular cleaning of the enclosure/dome is a part of their maintenance.

John, the requirement is pretty immediate (the wait is on our side). We were hoping for some plug n'play kind of setup for video enhancement. If not, we just go back to the drawing board and try to convince the client to setup a bunch of high power exhaust fans and/or one camera per display.

Matt, thanks... I didn't think of the refocusing problem. My understanding was that once the camera was setup (in a smoke free environment) it would come back to the same configuration regardless of environment being smoky. What a bummer!!!

In my experience, no PTZs I've ever used stores the focus point in the preset - position and zoom are stored, but the camera will try to re-focus anytime the zoom or direction are changed. There may be exceptions to this, but I haven't seen them. I'm thinking, too, that even if you did find a PTZ that you could store the focus and not have it attempt to re-focus, you're probably getting into a higher-end unit that's going to cost a lot more than it would have to simply have an individual camera for each panel.

And you still have the same problem that the further the camera is from the panel, the more smoke there is in between - optical zoom will get you a tighter shot, but the same amount of smoke will still be in the way. The only way around that is to get the camera physically closer to the panel... or find some way to reduce the smoke.

I don't think HIGH POWERED fans are required, and an exhaust fan wouldn't likely work too well, because as you draw smoky air away, it will only be replaced by more smoky air... the trick, if you go the fan route, would be to duct in some clean air from outside the affected area, and blow that GENTLY into the intervening space, to displace the smoky air. Using too much force would probably mean the fast-moving clear air would continue to pull smoke in behind it... a low, steady flow of clean air would be better.

Matt,

Strictly speaking, that's not true. PTZs can store focus settings (at least some can). If you turn off auto-focus on a Pelco Spectra and preset pan/tilt/zoom/focus, I'm pretty certain it will return to that preset (including focus). I'll check that capability on Monday and post the results here.

Matt, on a purge application you would ramp up with a variable speed 3phase motor different settings. This allows you to remove smoke at a different speed, triggered by a beam which allows you to see true activity. If you just fan a situation you create Flue Effect and just create a situation.

Removing the smoke slowly is like moving fog at a slower rate. This would allow you to see with a decent camera , but would not remove the particles left behind by the smoke and would be a problem later.

This can only be solved by using multiple items with a cameras to see the problem, a survey needs to be completed with all the environmental issues listed out and thought out.

There is no simple cheap solution to this problem. You need Thermal , UV , Slow smoke removal, and a maintenance schedule which cleans the lenses regularly.

Because Camera Design is set to open, close on light intake to the chipset, thru multiple chips. The imagers will only open, close on change. Smoke is like fog in that it has particles, liquid or gaseous, some solids.

Christopher, I get the impression you're not actually reading this whole discussion, or Pushkar's requirements.

They're trying to view the control display panels on various machines. Thermal won't show anything useful. UV filtering won't help because what they're looking at would seem to be lighted displays.

The entire environment is smoky; "removing" the smoke isn't feasible (if it were, I'm sure it would already be done). They only thing that MIGHT work is bringing in clean air to "replace" the smoke in small specific areas (ie. directly between the camera and panel), and admittedly, that's not guaranteed to be effective either.

-----

Carl: that's why I said, *in my experience*. I've never seen a PTZ that didn't autofocus after a move/zoom, but then I've never needed one to NOT do that.

The real trick, if it does store the focus setting, is whether it will do that while autofocus is on (so you can set your position/zoom/focus, then turn off AF), or if you have to turn off AF first and then manually set the focus before storing the preset.

I'd try it here, but I've been unable to get into the OSD on my cheap no-name offshore knock-off PTZ :)

I appreciate your response, and I understand what the client is looking for, but the only solution is to send either a display from each machine to a computer which in turn displays this on a screen which an operator can view all displays at one time or individually. Or send status pt alarms, troubles to a monitor which give s each machine, defines each condition and status and alerts the managers or operations of the problem.

Simple viewing wont solve the problem. Removing smoke at a slow rate will allow the system to be reviewed and not create a problem.

"Simple viewing wont solve the problem."

But... that's exactly what they're asking to do.

I agree with the second comment by Chris Freeman. Cameras really do nothing for their application and are probably more expensive than conventional analog signal transmission of the data.

Send an output signal that corresponds to the display they wish to see, to a local computer where smoke is not an issue. Either that or use Remote Desktop to view the local computer where the data is being sent.

If each device/signal is discreet, can the display provide "retransmission" of the signal? If so, then there are any number of devices that can take remote I/O and use various forms of communications to connect to another computer display. If not then changing to one that can might only be a couple of hundred max.

Pushkar, I checked with DefVision, and they say they can probably deliver what you need, based on the Lyyn website video. They have a 4 channel unit and they can probably analyze your recorded video to see if it's what you need. You can connect with them through www.defvision.com and the info@defvisionhd.com email will get you the technical assistance you need.

I took a look at Defvision. The product's spec look like a better match though it's pretty disconcerting that they have no phone number, nor physical address listed. How big / mature is this company?

Right. That'what I provided the email. The product originated in Japan and the 2012 formed US group co-owns the IP, and has sales, support and development responsibility for most of the remaining world market. I'll gladly connect you with the tech team, but figured you'd want to go thru the paces yourself. I am impressed by the box and am delivering it into 2 tough environments. I believe it can be a good alternative to thermals, in many cases, as it can be used to spot check locations for a period of time & then moved to another location. It can also be used as a post recording forensic tool. I will also keep a unit to offer customers such forensic analysis as a service. If a customer thinks they need a (or many) thermal, a difficult eval awaits and they often do nothing, or spend crazy $. To me, this is the eval tool to see if enhancement will do the trick and allow for a more confident, one-at-a-time thermal replacement or addition.

Also private domain registration and site is built on word press...

I spoke with DefVision recently and it has a number of limitations. Most importantly, it only accepts analog inputs. Therefore, it is limited to SD resolution and requires a decoder and encoder on either side, in addition to their processing box.

Their units support 4 channels fixed, with an MSRP of $12,900.

I obviously don't know the quality but it's going to be pricey and, if you are using IP cumbersome.