Advertisement

Blog

Calibrating for Product Integrity

This week I heard one of my favorite all-time words used in the context of having the proper reference to achieve a credible measurement. That word was “calibrate.” This word has so many connotations that have to do with accuracy, reliability, credibility, etc. The Merriam-Webster dictionary defines calibrate as:

To standardize (as a measuring instrument) by determining the deviation from a standard so as to ascertain the proper correction factors or to adjust precisely for a particular function to measure precisely; especially: to measure against a standard.

Every engineer knows that a design is only as good as the actual measurements can confirm. That being said, the integrity of the measurement depends upon not just the operator's ability to perform the measurement but also the integrity of the equipment used.

I was the 24th employee at what became a 1,200 person company. We were pioneers in the point-to-point digital microwave telecommunications industry. We had expanded to several buildings in a campus-type environment. The R&D engineering was separate from the manufacturing operations so electronic test equipment was spread out between two buildings as well. Customer service also had a dedicated lab for repairing returns. This scenario translates into a rather interesting asset management challenge. To make things even more interesting, 80 percent of the test equipment was leased or rented.

There was no single person assigned to manage the equipment. Each department rented their own gear as needed and when a company grew as fast as we had, the perspective was to keep pressing forward in terms of daily demands and priorities. As a result, some of the rented equipment was “lost” in time and space. When we tried to find the equipment to return it to the rental company, we could not locate it because it moved from lab to lab and into private offices without transport records. We were spending thousands of dollars on rental equipment each month and did not have the ability to justify the ongoing rental fees because we did not know if the equipment we were renting was still needed or not. So a day of playing “find the test gear” was mandated by management.

What we found was most disturbing: 70 percent of the equipment used in manufacturing operations was beyond the calibration date. We had no way of knowing that the measurements taken over the last X months were accurate. We found rental gear in cabinets that had not been used in two years because we were no longer making the products that required that equipment.

So, here is the best-practice when dealing with test equipment, rented or owned. As soon as the equipment arrives on site, a designated person assigned to asset management records the make and model number of the incoming gear. Also, the last calibration date is recorded and the name of the person responsible for the equipment is also noted in the record. The anticipated rental period is entered into a database cross referenced with the make and model, equipment location, and calibration date. The database includes a feature for real-time reminders indicating both calibration due dates and rental status payments.

Part of the procedure for managing the test equipment includes a travel log that details any movement of the gear from one lab to another or from one person to another. The borrowing lab or individual must sign a receipt for the equipment to be followed by the database update.

It is an excellent idea to have the rental company track the calibration dates as well. Setting this up requires an agreement with the rental firm that they will call a week in advance to allow time for any internal adjustments that would accommodate the absence of the equipment while it is out for calibration.

As a final comment, I would like to remind everyone that part of a quality audit should be the examination of all test equipment calibration dates. If you are using a contract manufacturer, by all means look at those dates. Ask the contractor how frequently the equipment needs calibration as stated by the manufacturer and if they have a calibration management system in place. If you are a manufacturer with test equipment, the only way you know your product is meeting spec is if your measurements are made with up-to-date calibrated equipment.

Don't run the risk of losing key customers or prospects by discovering too late that your data has not been reliable. Also, you may want to do a physical audit periodically to make sure that you are not paying for equipment that is not being used. Money is one thing. Design and performance integrity is everything.

7 comments on “Calibrating for Product Integrity

  1. t.alex
    November 27, 2012

    I understand a number of electronic products require good calibrated instruments during the development phase for things to work correctly. Products like cellphones or those for wireless communication, wifi, etc. depend a great deal on this. 

  2. dalexander
    November 27, 2012

    @t.alex…You are right. Any product that requires an active signal, digital or analog, output to the outside world or even just internally, is subject to being measured against a test specification. The only way to be certain that the specification has been met is to confirm the accuracy and up-to-date-ness of the calibration for the test gear. Mechanical measurements also need calibration as some are electronically based or even subject to wear and tear over time. Hand held digital or mechanical calipers should be zeroed before measurements. Are you old enough to remember the analog VOM that would be calibrated by holding the probes together and setting the needle to zero? I am not old enough to remember that. Yes I am. 

  3. prabhakar_deosthali
    November 28, 2012

    Yes Douglas , I am from that era of Analog when we used to short the probes of a VOM and calibrate it to show a Zero.

    While I was working in the controls lab of an Atomic test lab we had a predefined routine in which the schedule for calibrating each instrument was defined. A lot of these instruments were Analog .

    Later while working in an automobile company as IT manager , we had put locks in the system to lock loading of jobs on a particular machine if its tools were due for recalibration.

    As you say Calibration of instruments is a very important function in any production facility and a separate calibration lab and calibration staff is a must.

  4. dalexander
    November 28, 2012

    @prabhakar…what kind of equipment could you lock out users? Would that stop a line sometimes? Did you have the calibration done in house? We had to send spectrum analyzers and BER and the like back to the manufactures or use a dedicated outside calibration company that had been certified on each piece of equipment. It was not cheap and only got more expensive over time. How did your company manage calibrations?

  5. prabhakar_deosthali
    November 29, 2012

    Douglas, The kind of locks we had put in the software systems  for job loading on the machine tools in a automobile company. The system would keep track of the no of jobs done on a particular machine and generate alerts for calibration of tools whenever the set limit was about to be reached. If the limit was reached the software would block any loading of the jobs on that machine.

    Since the operators would totally rely on the computed generated job slips, the machines on which the tools were not recalibrate would remain idle  and the shop supervisors would immediately notice the same

    The calibration was done in house by the standards laboratory whic would also do the inspection of the finished jobs.

    As far as some electronic equipment was concerned it had to be sent to outside party for calibration.

  6. dalexander
    November 29, 2012

    @Prabhakar…I can bet that the calibration folks had to move quickly when it came time for the tune-ups. If your capital equipment cost several hundreds of thousands of dollars as in the case of numericaly controlled (NC), then it cost real money just to have any down time at all.

  7. t.alex
    December 2, 2012

    Douglas, unfortunately no I don't remember such VOM 🙂

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.