Articles

FROM CONVENTIONAL TO FREE FORM PRODUCTION

Pitfalls for traditional typically smaller labs. Technology suitable for Freeform production has become available even for smaller labs during the last few years. Due to limited resources special problems arise for those labs, especially in lens quality ans process control.
 

QUALITY OF FREEFORM LENSES

During the last 25 years FreeForm  production developed  from prototypes for large labs to a standard  technology  even  suitable  for  small labs. Today nearly all invest in new machinery is done  in  this  new  technology.  Also  nearly all  new  lens designs entering  the market  are FreeForm  designs. Smaller labs that are stiII producing conventional lenses are faced with the demand to switch to FreeForm  production sooner or later.

The quality of FreeForm lenses consists  of three dimensions:

I   Manufacturing quality

I   Lens design  performance

I   Quality of refraction  data

The manufacturing quality includes all steps in the production of a lens. The production process of a FreeForm lens is much more complex than the conventional production, which means there are much more things that can go wrong. Therefore more know-how is needed to reach the same level of production quality in FreeForm production compared to conventional production.

The lens design performance describes the quality of the eye model used during calculation of the lens. The range of available designs in the market is huge. The first designs that were available were more or less conventional progressive designs that were produced with the progressive surface on the back side instead of the front side. These designs are very similar to the corresponding conventional designs. Then the calculation program designer started to develop more optimized and later on individualized designs, which brought the lens design performance far beyond the possibilities of  conventional designs.

The perfomance of a lens design grows with the know-how invested in R&R to develop the design. Standard designs are conventional designs taht are being produced on the back surface, without any optimizations and individualizations. The next step leads to optimized designs that are using ray trackings or similar technologies to get better lenses. Individualized designs take custom specific parameters into account to futher optimize the lens. 

But there is an opposite development as well: Some designs in the market do not even have the performance of conventional designs. It was costly and therefore risky to develop a new progressive conventional design, so care was taken to establish a high quality. To develop a new FreeForm design calculation program, even with a lot of individual parameters, is quite easy. But the most important part - the eye model - is sometimes very poor and therefore the designs do not perform well.

The third dimension is the quality of the refraction data. With individualized FreeForm designs the amount of parameters that need to be measured by the ECP grows. In some markets in the world it is still common that the ECP changes the refracted data to get lenses that thinner or can be mounted in the selected frame. The total  quality  that the end  customer gets is  the minimum of all three dimensions. That is only one of these quality dimensions is poor, the  produced lenses will be poor as well, no matter if the other dimensions are perfect  or  not.

These  quality  dimensions  are  the  same for larger and  smaller  labs,  of  course.  The difference is made by the market segments smaller labs are addressing, and the resources and knowledge they have. 

SMALL LABS

Smaller labs are addressing a different market segment than large labs do. Typically smaller labs are specialized to individual customer solutions. These solutions are often  enough the only reason why they exist, especially when large labs or companies are present in the  same  market.

The limited  resources  small labs have  are normally concentrated in marketing, sales and production.  They  do  not  have  own   IT if departments, engineering or even R&D. So the knowledge needed for FreeForm production in these areas must be bought from the market. A small lab therefore has the problem that it needs to find partners that not only help to set up a basic FreeForm production, but a FreeForm production that has the ability to continue the individual customer solutions. Further to this, the partners must be able and willing to transfer know-how not only to start the FreeForm  production, but also during production, to keep it running and to further develop the lab. On the other hand the small lab must understand that it will be dependent on th is, and that it has to  be  called  'help' and  not  'intervention'.

Two of the main services from those partners must be help to detect and eliminate failures in the production and help to establish a process control.

LARGER LABS WITH FMEA PROCESS

Every large lab with its own engineering or even R&D department has developed systems - often together with the IT - to set up its own FMEA process. This way they have a guideline in case of a failure, which possible causes are likely, how  to  check  for  the  real  reason, and fi nally how to get back to normal operation.

Even in large labs it takes years to set up such a knowledge database. It is impossible for a small lab to get this work done by the lab itself. They simply do not have the resources, the machine capacity and the time to do all the needed tests and documentation on their own.

On the other hand it is difficult for the larger labs to provide their knowledge database in a way that small labs can easily query. The knowledge databases of the large labs typically contain much more facts than necessary for the small labs, but an automatic filtering that only presents the needed data to the small labs is  difficult.

One way to come around this is an engineering service provided by the partner that receives the failure description, and from there controls the process  to find the cause and eliminate  it.

On the one hand it queries the large knowledge database and filters the information to match the needs of the small lab, and on the other hand it can tell the small lab, which additional checks are needed or what needs to be  done. It is obvious that this FMEA process needs a good cooperation between the partners. But still with a very good cooperation, this process will take a bit longer to find a solution for a failure than an in-house FMEA process would take. This fact combined with the knowledge that in a FreeForm production much more can go wrong than in conventional  production, the need  of a process  control becomes  clear.

PROCESS  CONTROL

A process control helps to detect upcoming problems before they lead to rework and breakage. In conventional production, process control hardly happens in small labs. The employees are standing next to the machine and   'feel',  if  something   is  not  working correctly. This kind of feeling does no longer work in FreeForm production, since too many things can go wrong, and most of them are neither to be seen nor to be heard. The idea behind the process control is to inspect sample lenses after each production step. These sample lenses can be either test lenses or lenses from the running production, it depends on questions like 'can those len ses be delayed' or 'is  the planned  test  a  destructi ve test'.

The process control differs from normal lens inspection, which is an absolute requirement for every lab and every produced lens for both conventional and FreeForm, in many things. The main difference is the point of view. While the lens inspection judges, if the lens is okay or not, the process control targets your production steps. For each production step different quality criteria are defined. There are bou ndaries for 'totally okay', 'take care', 'react now ' and 'out of param eters, stop  production.

While a production step is in status 'take care', one or more parameters started to shift.  At that moment the quality of the final lens will still be okay, but it is clear that if the shift continues, some action must be taken. So you have time to schedule are calibration or maintenance.  Since  many  small  labs  are not located in industrialized countries next to an international airport, it can take days or weeks until this is possible, so it is important to detect those  shifts as early  as possible. If a production  step  reaches  'react  now'  it  is  very  close to a failure and immediate reaction is needed , otherwise rework and breakage will  the result. Since small labs often enough  small  a single line of machines, a failure in one production  step  means  that  the  lab  is  out    of business until the failure is eliminated. Therefore  as many potential failures as possible have to be detected by process control, before  they  grow  to failures.

CONCLUSION

When it comes to invest to new mach i nery an FreeForm technology is the only choice, even for smaller labs. But it will be a huge step for small labs, which they  cannot  take on  their own.   They   will   need   a  good   network  of partners, consisting of  machine  vendors,  lens design  vendors  and  lab management software providers. Together  with  these  partners  small  labs  have  to  find  a  solution  for their special market  situation.

The partners  must support in managing the switch to FreeForm in keeping the production up and running in analyzing failures, in establishing a process control ans training employees and opticians.

Thomas Mischke studied Industrial Engineering at the Technical University of Darmstadt. From 2001 to 2007 he was Backend Software Developer at Interactive Data Managed Solutions in Frankrurt. One of the leading companies providing software and services to financial institutes. In 2007 Thomas Mischke joined optoVision. A company of the Rodenstock group. He is responsible for t he overall design and development of the lab management system (LMS) ‘optoWare’. its integration a nd operation not only in-house at optoVision but further in customer labs worldwide.