Tracking Down Lab Breakage & Errors

By Linda Little
A s any good detective knows, the clues to solving a crime will lie in the knowledge of what is constant or ‘normal’ and what is different. It is the ‘different’ that begins the process of identifying the clues, challenging and inevitably, solving the crime. Detectives follow this formula over and over. When it comes to lens manufacturing processes, though, tracking lab breakage is considered much more complicated, or is it?

Let’s take an example of a lens that is at final inspection and the axis measurement is off. This would be considered an error or breakage since the lens can not be used and must be replaced. This is a substantial cost for a lab. Where would a detective begin?

Most likely, a review of the steps to process the lens would be in order. Identifying the ‘normal.’ Let’s look at this through a diagram for ease of explanation and reducing the boring reading! So continuing with an Off-Axis Lens as our example, the Detective will need to realize the Inputs to the Process and the Outputs or often referred to as “Requirements” of the process. “The first step is to measure the error,” explained Jack Banville, VSP Labs. “We want to know if we have a problem or not and the size of the problem.” At VSP, employees are trained to use a system of ‘Statistical Process Control’ when handling a breakage. “We started using the Quality Improvement Process by Crosby, in 1992,” Banville added. “Since implementation, except for 1 or 2 years, our breakage rate has decreased every year.”

What does this system offer the Detective? “We don’t want to solve the same problems over and over,” he explained. “We document the solution, document the procedure and don’t change it.” This takes a great deal of discipline to make sure the procedures are followed. VSP measures their output to make sure the processes are correct and to prevent breakage. Ah, that’s the clue. Prevent the breakage to begin with. Is it possible?

Within the context of the Quality Improvement Process, Prevention is causing something not to happen, or conversely, making sure that something does happen exactly as it should, every time it should, with consistent outcomes. Ideally, this leads to zero breakage every day! So why, with Quality Improvement over 25 years old in the United States, are we still talking about tracking down lab breakage and errors?

Let’s go back to the Inputs column in our chart below. Each input has to perform exactly as it is expected to, based on the requirements for the job. For example, the people must be trained to operate the equipment, follow the procedures without tweaking, understand and use the lab management software without compensating, choose the correct consumables, and know how the lens products perform under different conditions.

The equipment must perform as expected, the procedures must be up-to-date, accurate and proven through measurement, the lab management software must be current with correct lens data supplied by lens suppliers, up-to-date with materials, calculations, trace information, etc., the consumables perform as expected and the lens material performs as expected. I guess if you cross your fingers, close your eyes, and bite your tongue, you can hope each step is in sync, because to meet this level of performance, each item, each step has a number of inputs that could affect its performance.

If we go back to our lens that is off axis, the first step back would be to determine how many lenses have been off axis, and when they were processed. Narrowing down the number of errors, when they occurred, any similarities of errors – same equipment used, same operator, same lens type… “There are repeatable errors,” explained Chris Bowers, operations for Walman Optical.

“We look at our measurements of breakage and identify the top three biggest cost reasons for spoilage and eliminate them.” Walman’s measurements are based on ‘out of ordinary’ work flow. The job is stopped until a solution is found. “Usually the primary causes are calibration or human error,” Bowers added. “We know how to process lenses, the variations come when the lens isn’t what we think it is, the machine doesn’t perform the way we think it should and the process isn’t followed the way we want it to be.”

A simple example he sites, “We know that we need to change the blades at 300 cuts. Someone decides that we can save money, and that the blades can run another 100 cuts. The prevention is gone and some portion of those 100 cuts will result in spoilage.” People are natural problem solvers; we all have spent years performing creative solutions to routine problems. When job descriptions are written, often these are defined as job tasks or responsibilities. We know what the end result should be, so we tweak and maneuver the ‘procedure’ to produce the result.

Our poor detective is somewhat stumped at this point with all the possible clues to follow, the bad lens could be a result of a calibration error on any of the machines, slippage from a consumable, tracing, or the coating brand. Where to begin? According to Steve Dombey, with CC Systems, one of the most common errors he sees is in the area of Protocol and Procedure.

“The lab management system produces a work ticket and experienced operators are relied on to make the changes along the process to tweak the information,” Dombey explained. “The job ticket should be accurate to begin with.” Each lab is unique in its operations and procedures, according to Dombey. Therefore, the lab management software would need updating to meet the lab’s requirements for processing different materials, different prescriptions, or perhaps want additional or different information. “The software should be updated each time a change is found so no one has to ‘work around’ the job ticket,” Dombey added. “The discipline of identifying the solution and correcting the problem permanently is the hardest to implement.”

Dombey is an advocate of measuring the problem and working with the operators to use their knowledge and implement it into the software as the permanent solution. Then operators don’t have to remember to tweak and manipulate to meet job requirements.

The lab manager can easily adjust many lab management software products, while others may take more effort and partnership with the supplier. The point is right though, and echoed by Banville and Bowers, don’t solve the same problem over and over. It is too costly – from the time it takes to correct, the actual material costs, the operator time and the customer frustration, stopping to correct the problem permanently seems like a good idea.

Perhaps inspection is the step to prevent a bad product from shipping to the customer, but it certainly isn’t a cost effective method. As our Detective works backward from Inspection, each step is scrutinized for variations. Is the edger calibrated correctly, is the blocker on axis, was the finish pad aggressive enough for the coating, does the pressure need to be increased to minimize torque, is the problem off one edger or all edgers, and on and on until the differences from normal can be identified.

“When it comes to machinery, one has to have an open mind,” commented John Stellano, technical support representative for National Optronics. “Anything can cause a problem. Jumping to solutions is like chasing your own tail, you won’t get anywhere and the customer will be frustrated.” If the customer has been tracking the problem and can isolate what’s different, solutions can be determined and corrections made efficiently and effectively get the operation back on track, according to Stellano.

When breakage occurs, think like a detective to look for clues. A good process, protocol and or procedure should be in place to prevent breakage to begin with. When it happens, and it will, the clues will be there through measurement to identify what’s out of sync. Once a solution or solutions are identified, implemented and measured for success, re-define the procedure for all to use.

It is human nature to solve problems and move on, the real discipline in tracking breakage in a lab, is to encourage human nature to only solve the problem once and eliminate the opportunity for it to happen again. Now if we could just get the machines to adjust themselves, the lens material to be consistent in all conditions and consumables and coatings to perform as expected every time, zero breakages might be possible. Until then, tracking breakages offers a great opportunity to build prevention into the lens process and consequently reduce the costs of breakage in the lab. The information is valuable for identifying where and how the variation is occurring, identifying solutions, and eliminating the problem.

Solving problems shouldn’t be any more complicated than knowing what is constant and what is different, unless of course, there isn’t a clear, repeatable, “constant” to work with.


CURRENT ISSUE


Labtalk September/October 2018