Thursday, January 6, 2011

What's the Difference Between an Optical 3D Measurement System and a CMM? (or, why optical 3D metrology suppliers have a bright future but grey hair)

One of the reasons we are blogging about 3D metrology is because it's an emerging field, with many companies offering a diverse array of products to suit a wide range of customer needs. It is by no means a new field, but it seems that even those who have worked with 3D metrology for their entire careers are interested in how an optical 3D metrology system differs from a CMM.

Fundamental System Differences

To a CMM, every part looks the same, provided it is rigid (i.e the surface is "solid"), and the part is not moving (i.e. it is fixtured to the CMM's table). Metal? Carbon fiber? Hard plastic? Transparent glass? No problem. Lights on in the room or off in the room? No problem. Polished surfaces? No problem. Therefore, the system's usage is dominated by technique, provided that the CMM is accurate enough for the job, etc., etc., etc.

To an optical system, surface finish properties have severe consequences in relation to the results. Metal versus carbon fiber? Potentially big differences, depending on the surface finish. Hard plastic versus metal? Potentially big differences depending on the surface finish. Transparent glass versus hard plastic? Potentially big differences depending on the surface finish. Ambient lighting changes? Potentially big differences. On the other hand, an optical system doesn't care whether the part is rigid, fixtured, or even sitting on a table at all. Therefore, it excels at rubber, paper, harsh field environments, fabrics, generally moveable objects, and objects that don't lend themselves to sitting on a surface plate.

The primary difference between a CMM and an optical system, then? The CMM's part measurement category is fixed (i.e. rigid/fixed/fixtured), and each measured point is a function of operator technique. On an optical system, each measured point is operator-independent, but whether the measurement is successful or not is dependent on the technique of the system design. Can an optical system be designed to measure ONLY glass? Probably. ONLY shiny metal? You bet. ONLY carbon fiber? Of course. Therefore, it is the optical system that determines the success of the measurement. Using the wrong optical system for the job results in disaster. To have an optical system that can measure glass AND shiny metal requires two separate systems, combined. Get it?

A Bold Assertion
The primary reason that many optical systems use "spray," "powder," "paint," or "coating" as part of their process is an attempt to make all of the surfaces "look" the same to the measurement system. This way, whatever part is being measured, as long as it's a matte white finish (or other controlled finish depending on the optical system), the system will always "work." This is great for a laboratory that can take time and make special preparations for each part being measured (and in many cases, contaminate the part), but in production, it's a non-starter.

We're going to make a bold statement here: you should not spray, powder, paint, or otherwise coat the surface of an object prior to measuring it optically. Customers (i.e. you) hate it, and therefore it will forever be banished to the history books along the ranks of car seats that didn't have headrests, tubed tires on cars, and steam as a fuel source for cars.

Not only is "spray," "powder," "paint," or "coating" messy, time-consuming, and absolutely ridiculous when you think about it, but it absolutely kills your measurement system's accuracy. If you don't care about accuracy, then spray all you want, but many people do, and those are the ones who should be reading this blog.

Anyway, back to the topic...

Mistakes Companies Make
Many manufacturing companies that have worked with optical metrology as a potential CMM replacement have spent millions of dollars testing, researching, and trying, trying, trying to make them work side-by-side with a CMM. Some companies have been trying for decades. Some of them have been successful, in fact (the growth of optical 3D systems has led some CMM companies to also offer optical technologies). Of the manufacturing companies that have found success, few, if any, however, found the right optical technology on the first attempt. In most cases, a wide range of technologies were purchased, tested, and developed, and one or two may have been found that are "good enough."

But, for every company that has succeeded in making optical technology "work," many still do not, to this day, have a solution. Why is this? In some industries, optics are used to measure nanometers. In other industries, optics are used to measure galaxies. In others, optics are used to measure non-visible properties. Why can't every customer have a solution? Do you see where we're going with this?

If you haven't found success in your optical 3D metrology, we assert that 9 times out of 10, it is due to system design. The system design relates to the total sum of trade-offs applied to each individual customer.

Mistakes 3D Metrology Companies Make
The primary mistake that 3D metrology companies make is that they try to sell ONLY a pre-packaged "product." While this makes sense for some customers, the remaining customers are left out in the cold. The ones that are left out in the cold are usually manufacturing companies that want to apply optical 3D metrology in production. The "pre-packaged, off-the shelf 3D metrology products" don't cut it, haven't cut it, and never will cut it in a production environment for some customers. Why don't the 3D metrology companies, then, make customized 3D measurement products for the manufacturers who are currently left out in the cold?

Answer: because it's hard. Most 3D metrology companies prefer to design a single "general-purpose" system, and sell it to as many customers as possible. Creating an individualized system for each customer does not scale well for many companies (it does scale well for Twin Coast Metrology, however. It is, in fact one of the primary aspects of our business model, and is also why we are able to GUARANTEE each system we install).

The manufacturers making production parts often have very, very difficult requirements. They are, perhaps the top 5% of the industry in terms of difficulty. Yet, they also probably represent 50% of the marketplace in terms of market share. The 50%, of course, is divided between hundreds of unique applications (measuring glass, measuring shiny metal accurately, measuring strange surfaces, and quickly, in robust ways, without spray/coating, with a single push-button). Every single customer's system in this top 5% is a grueling, challenging battle, and will produce many grey hairs on the heads of the 3D metrology companies.

This, of course, is much better than producing many grey hairs on your (the customer's) head over decades and millions of dollars.

A Caveat
There has been a push in the 3D marketplace by "general purpose" optical systems that can supposedly measure a variety of surfaces that have traditionally been sprayed/coated/powdered/painted, wihout the use of spray/coating/powder/paint. Tread very carefully on these waters. It is very difficult to switch to a non-spray approach. Again, it is hard to design a system that will measure a range of surfaces, and it often requires combining several measurement approaches. In many/most cases, the system has to be designed from the ground up in order to meet customer requirements.

Be very careful of any optical 3D metrology system that can supposedly measure a wide range of surfaces under a wide range of conditions.

Make sure it's your 3D metrology system supplier's hair that's grey -- not yours. Ask them to GUARANTEE the results, ask them to utilize custom algorithms and hardware, and perform the measurements without coating/spray. Then, and only then, will the optical system meet the CMM's capabilies on EVERY part - not just a limited few.