Does "Imaging Inspection" Have a Conscience?
ovitRecently, we have been dropping a large number of demo units, and we have been receiving a variety of evaluation results.
Some of them have been evaluated in comparison with those sold as "inspection equipment for XX" with handling. Although we cannot make simple comparisons because this is a "complete device" and this is a "simple experimental environment," we have generally received higher scores for the FI. However, the FI is generally preferred over the FI, although the FI does not get a perfect score of 100 points, with the comment that "some work needs to be done on the lighting.
What I was concerned about was that there were too many "x's" on the other side, even though it was a "completed device. I had checked the actual equipment in the technical report and at the exhibition, and I had a feeling that there was something fishy, but I was not surprised.
There are far too many stories of this kind.
Do they really want to "satisfy the field"?
Certainly, there is a high need for "visual inspection equipment," and it may be attractive as a business.
However, we cannot tolerate the attitude of proudly selling half-baked equipment.
Our stance is,
(1) Make one that is easy to operate.
(2) Make products with the strongest detection capability.
(3) Provide the product at a price range that allows a "trial purchase" in order for the customer to make a thorough evaluation of the product.
This is what we are trying to achieve. In any case, we do not want to make a "non-working visual inspection system.
Many of those in the "visual inspection equipment" business are reluctant to admit their own guilt. Most of them blame the workpiece. They may be a good money loser, demanding additional costs over and over again.
Are suppliers of inspection equipment treated as criminals? This is a harsh comment. It is often technically difficult to distinguish between good and defective products with the same concentration difference when viewed with a camera. We explain this to the quality manager (who has the right to purchase the product) prior to purchase, and while we are further priced out of the market, we often cut back on features. However, after the actual introduction of the system, the on-site operator often complains about the inspection results and we often have a dispute. We then have to charge additional fees and meet tight deadlines, but I wonder if many users think that they are being used as a money loser. It is a sad position to be in because our efforts are not rewarded.
I myself used to do the same thing as a "hands-on" employee of an inspection equipment manufacturer, so I understand how you feel. I became independent because I couldn't stand the thought, "Who will be pleased by this? I am now doing it in a small way (it is not a viable business if done properly), saying, "Let's make something that will make the field happy.
The problem is that "imaging inspections" have become so common that the "negative aspects" have not been exposed, so the "authorities" think that "if we don't introduce it, we will lose the competition" and "we can manage (even if we are faced with difficulties)" and "GO" with it. (This is becoming less and less common...)
And the ones who cry are the "field" and "the actual work force of inspection equipment manufacturers.
Image inspection" is suspicious. So
(1) Inspection equipment manufacturers should prepare an environment where they can evaluate "actual equipment.
(2) Users must evaluate the actual equipment to determine the degree to which it is feasible.
I think we have to create a "common sense" that "the degree to which it is feasible" must be determined.
We are now considering the introduction of a visual inspection system, and we have discovered that there is a certain "pattern".
The "pattern" is that we first send a "defective" sample to the manufacturer and ask, "Can you perform this inspection?" Most manufacturers respond, "Yes, it is possible.
Then, a screen showing the extraction of defective parts of the sample workpiece is attached.
However, the extraction method and lighting of the defects are often changed each time, not under the same conditions.
The manufacturer's verification is done under the best conditions.
If you do not have a clear understanding of this, you will often have problems when you install the system.
I am afraid to introduce a product without first seeing it with my own eyes.
If a demo machine is available, I ask to borrow one and touch it to see for myself.
If there is no lending service, I visit the manufacturer and check the settings and operability with my own hands.
This often yields results that differ from those in the manufacturer's verification data.
Maniacal settings cannot be used for mass production, so it is difficult to operate unless the settings are simple and the lighting and other settings are versatile.
We recommend that you verify the settings by yourself, and use the manufacturer's verification only as a reference.
Also, image processing is not a panacea, so it is difficult to replace 100% of visual inspection.
I think the way to "make the most of the inspection equipment" is to fully verify what can and cannot be done, and to make effective use of it within the "possible" range.
As Mr. Yamada says, "If we make it an "authority", we can introduce it,
If you are the "authority," you may think, "If I don't introduce the equipment, I will lose the competition," or "I can handle it (even if I encounter difficulties)," and "GO.
Our company's director also thinks the same way, and says, "First, install the inspection equipment! First of all, install an inspection system!
However, if there is a problem after introducing the equipment without much verification
If there is a problem, the inspection system is useless.
and that is the end of the inspection equipment.
If it is not usable, there will be no future development, so I think that sufficient consideration is necessary before introducing the equipment.
By the way, we are borrowing FI's demo machine, and it is amazing.
I can see a "light" in the introduction of inspection equipment.
Thank you all for your comments.
As #comments indicate, masa's place is currently undergoing evaluation with a loaner unit. They are comparing with several competitors.
Currently, our biggest competitor? is the preconceived notion that "images are no good" by users who have experienced "images". If this trend continues, I think there is a strong possibility that this industry will not be able to stand.
I think we need to rethink once again what we need to do to regain the trust of users.
I found this place after doing some research.
I am a sales representative for a manufacturer of inspection equipment using image processing.
I have looked at your site. I feel that FI is a very good product without actually seeing it.
I always consider it important because there will be no next inspection equipment if the field is not satisfied with it.
However, there is a growing tendency for the content required of image processing equipment that includes handling to deviate from the original purpose of introduction.
The most important thing is the identification of good and defective items, but added value is required for various reasons. For example, identification of defective items, visual inspection must be completely eliminated, etc.
In my case, I explain all concerns. Sometimes I say I can't do it, even though I'm in the imaging equipment business (and I often get credit for that). (Often I gain credibility by doing so.)
One of the cases where I get into trouble is that no matter how well I explain, there are at least some people who think that since they have made an expensive purchase, it is OK to add or change a few things later.
It is indeed tough to get work done later on an order where defects have to be identified item by item.
I know that there are many companies that do not, although all the opinions written about imaging equipment other than FI are very harsh.
By the way, do they ever hide the fact that they changed the lighting conditions or extraction conditions to suit their needs? I am surprised because I can see that it would be a big problem even if they receive an order. It is not a bad thing to change the conditions, but it is natural to include an explanation or description that "this example had to be done under these conditions.
It is true that some people have a preconceived notion that images are bad.
I think the problem with image processing equipment is that the focus is only on new technologies that have become available. Efficient operation methods? The concept of image processing? I think it is necessary for the further development of the industry to be generally recognized in the areas of efficient operation methods and the concept of image processing.
appleさん、コメントありがとうございます。
>サイトを拝見しました。実際に見なくてもFIは非常に良い物と感じます。
言葉だけで信じてはいけませんよ、この業界。
>現場が満足してくれない検査装置は、次がありませんので常に重要視しております。
たしかにこれが最重要だと思います。
ところがこれが非常に難しいのも事実。
>ただ、ハンドリング込みの画像処理装置に求められる内容が、本来の導入目的から脱線する傾向が多くなっています。
>一番大事なのは良品と不良品の判別なのですが、様々な事情で付加価値を求められます。例えば、不良項目の識別、完全に目視が無くならなければ駄目など。。。
「画像検査で何でもできる」といって売ってきた業界のミスリードの結果だと思います。
良品と不良品の判別だけであればそれほど難しい話ではないのですが、対象物によって異なる不良項目の識別はとてつもなく難しい話です。文字認識だ顔認識だと一生懸命に研究されているにもかかわらず100％は程遠い状況であることを考えると、一つの欠陥を認識することでも難しいことだと気づくべきなのですが。
あと完全に目視をなくすというのも難しいです。必ず「良否の判断が難しいもの」が存在し、それを画像で判断してあやまって流出させたら大問題。よって、機械的に「良品」と判断できるものを選別し、そのほかは人間の判断に頼るしかないと思います。
>私の場合、懸念事項は全て説明します。画像処理装置の営業なのに出来ないばかり言ってる事もあります。（それで信用を得る事もしばしば）
たしかに「できない」といえる営業が信頼を得るのは事実です。それぐらい業界は荒みつつあるのかもしれません。
＃でも「営業」という職務上「売らなければいけない」つらさもお察しします。
>揉めるケースの一つは、どれほどしっかり説明しても、高額な買い物したのだから、多少は後から何を追加、変更してもOKだと思っている方が、少なくともいる事です。
この答えが「完成品を売る」ことだと思います。
カスタムで受注生産をやっている以上、上記の問題からは逃げられないと思います。
>不良を項目ごとに識別しなければならないようなオーダーで、後からワークが出てくるのは、さすがに厳しいです。
これは絶対に無理です。
>FI以外の画像処理装置は、厳しい意見ばかり書かれていますが、そうでない会社も多いと思います。
あるとは思いますが、多くはないと思います。
つらいのは「正直にやれば商売にならない」仕事であるということ。これは間違いないと思います。
で「FI以外の?」ですが、現場を満足させられるものであればOKだと思います。ただ私自身FIに教えられたことは非常に多く、この業界のこの状況の中で「伝えるべきものは伝えなければ」と考えています。判断は「現物」でやって貰えればよいと思います。
>ちなみに照明条件や、抽出条件を都合に合せて変えた事を隠すことってあるのでしょうか。たとえ受注になっても大問題になるのが見えていまるので驚きです。変える事自体が悪い事でなく「この例はこの条件でないと駄目でした」の説明か記述は当然入れます。
appleさんのような営業さんばかりなら、今の状況にはなっていないと思います。他社との比較結果を見せていただくことも多いのですが「オイオイ」というものが多いです。だからmasaさんの「自分で確認する」という姿勢は非常に重要です。
>たしかに一部で「画像はダメ」という先入観を持っています。
>画像処理装置は、新しく出来るようになった技術ばかりが注目される事に問題があると私は思います。効率の良い運用方法？、画像処理の考え方？の部分が一般的に認知される事が業界のさらなる発展に必要と思います。
一緒ですね。
私はたまたまFIの方法を見つけてしまった。ここから得られた思想を伝えていかなくてはならないと考えています。
http://www.imaging-association-japan.com/seminer.html
でタイトルを見るだけでも当方のものはどうみても異常。
「現場」を考えるとこうであってはいけないんですがね。
＃もっとも「確信犯」なんですが・・・。
appleさんのような業界人が増えることを期待します。