Infrared as a service

[Mikkopee]

Hi all,

We made some research of using infrared in Finnish utilities.
Based for results and some assumptions, we made some monte carlo simulation and analysis of results whether to buy lowcost imager, high-end imager or use service.

If You are interested on this subject, it is available at

http://www.uwasa.fi/~f76331/infrared.html

I am very thankful, if You can comment the paper.

Cheers,
Mikko

 

[IRstuff]

The actual link to your paper is:

http://lipas.uwasa.fi/~f76331/13E.pdf

> Obviously this is not written by an English speaker. I suggest that you find someone fluent in English and English idioms to edit your paper.

> Usage of "measurement units" instead of pixels is highly non-standard.

> Assumptions about IFOV appear to be baseless, since IFOV is a function of the lens and a smaller FPA can be supplied with a longer focal length lens to achieve the same spatial resolution.

> You make claims about the apparent sensitivity of the cameras as the basis of the fault detection probabilities with no substantiation, other than the supposition that the smaller array is poorer.

> In general, the ability of a camera, given sufficient sensitivity and resolution, to "find" a fault is a function of the quality of the operator. Therefore, your premises and conclusions are not substantiated.

> The basic conclusion seems to be overblown. Your premise that the expensive camera can detect 100% of the faults will automatically lead to the final conclusion of "more expensive is better." The Monte Carlo analysis for comparison seems to be pointless, since you are using probabilities and assumptions that appear to have no real world basis and that automatically skew your comclusions.

TTFN

 

[Mikkopee]

I am sorry for not being fluent in English. It is always hard to write in foreign language, hopefully text is still understandable.

Few remarks on Your comments:
Of course operator skills and camera accessories (lenses etc) have significant effect. We tried to make some difference between expensive and low-cost cameras and we thought that best way to analyse the difference is size of array? So that sufficient resolution and sensitivity is missing from the low-cost cameras on some situations (long distances when imaging high voltage equipment).

We assume that 100 % of these faults will be found with expensive camera, since they were found with it.

Final conclusion is that for a single utility service should be chosen. Single utility should also buy a low-end camera instead of high-end camera when comparing only economic reasons. So we are not stating that "more expensive is better". Monte Carlo analysis is done to smoothen our errors in assumptions, but of course, probabilities we have chosen have significant importance on the results.

I am extremely thankful for Your comments, TTFN. I will review the paper with Your comments. I would also like to know, if someone has tested the difference between low-end and high-end cameras to get more "real world basis".

Best regards,
Mikko

 

[IRstuff]

I understood what you had written in the paper about the expensive camera. My issue was with the a priori conclusion that the cheaper camera would not find 100% of the faults. Many of your conclusions in paragraph 4.1 are tied directly to that initial presumption.

TTFN

 

[IRJohn]

As I've said elsewhere on this messageboard, I suggest we make a distinction between "detection" and "analysis."

Given that,160x120 systems can provide excellent detection (assuming conditions are right) and some, perhaps limited analysis with a standard lens; when it is possible to move closer or add a telephoto, even these smaller detectors can provide excllent analysis capability (but with a smaller field of view).

I have side by side images taken with two different cameras, a 160x120 and a 320x240. Both were focused and "auto" adusted in as fair a trial as is possible. Both are very acceptable in my opinion. I would be happy to share these with anyone who would like to see them

We are seeing considerable pressure on the market to buy bigger and more expensive cameras. For some work, they are useful and for other work, they may be essential. In many instances, a good argument can be made for using BOTH levels of camera.

As an analogy, if I need an SUV, I should buy one, but if I can get by with a smaller vehicle, why not. In either case driving too fast for conditions is not safe! The same can be said for ANY camera in the hands of someone who is not trained to use them or who fails to understand the impact of ambient and system influences.





If I can be of further assistance, please feel free to contact me.
John Snell
Snell Infrared

http://www.snellinfrared.com

 

[eyec]

good points here, but i see the basic premise of in-house vs contracted IR service as one of cost-benefit.

all cameras being equal - the technician is the deciding factor, with their experience and familiarity with the equipment used and that being inspected.

if a company has enough IR surveys/inspections that will keep the technician (productively) busy at the same time allow them to stay current with the technology, then it may be more cost beneficial to perform their own IR services.

on the other hand if they only perform intermittent "scans" for problems, it is my opinion that they will cause more problems/concerns with possible false indications or questionable results. therefore contracted services would be preferred.

finally, stand alone IR for any mainteance tool is only a small part of a predictive maintenance program and should be considered in total - IR, oil analysis, vibration, acoustics, etc.

this may give more credence to performing in-house IR/PdM than using contract services. absent a PdM program i would recomend contract services every time for their full-time expertise instead of a part-time in-house technician. regardless the cost of the camera!