Text preview for : 5992-0037EN Assurance of Calibration Results_ Applying the Power of Visual Information to Improve Qu part of Agilent 5992-0037EN Assurance of Calibration Results Applying the Power of Visual Information to Improve Qu Agilent 5992-0037EN Assurance of Calibration Results_ Applying the Power of Visual Information to Improve Quality c20140908 [11].pdf
Back to : 5992-0037EN Assurance of | Home
Keysight
Assurance of Calibration Results:
Applying the Power of Visual
Information to Improve Quality
Application
Note
Abstract
To help you mitigate the risks inherent in your measurements, Keysight
maintains the ongoing reliability and validity of our calibration processes.
We start with systematic monitoring of results from our highest-volume
instrument calibrations, as performed using automated software at any of
our calibration sites around the world. Gathering data is just the starting
point. Turning data into information is the next step, and visual informa-
tion--in the form of control charts--drives decisions about the need for
improvement activities and corrective actions. Three success stories show
how Keysight ensures the quality of test and calibration results.
Introduction
During the development and manufacturing of your company's products, the speciied accuracy of
test equipment helps you make key decisions: Is a design meeting its performance goals? Do the
products coming off the manufacturing line meet their speciications? Are they satisfactory to ship?
Our calibration services help you answer these questions about pass/fail decisions by mitigating the
risks inherent in the measurement process. One important way to do this is by actively monitoring the
ongoing reliability and validity of our calibration processes.
We do this through systematic monitoring of results from our highest-volume instrument calibrations,
as performed using automated software at any of our calibration sites around the world. We save
the individual points for every performance test. All this data is collected in a database that contains
billions of data points. Collection happens automatically in the background and has no effect on cali-
bration time.
Gathering data is just the starting point. Turning data into information is the next step, and informa-
tion drives decisions about the need for improvement activities and corrective actions. Going from
data to information to action requires processes, tools and people. Two standards guide our process:
ISO/IEC 17025 and ANSI/NCSL Z540.3. Two essential tools support the process and serve the people:
graphical presentation of data and our own Calibration Measurement Monitoring System (CMMS).
The key people engaged in the process are quality assurance engineers, service technicians and,
when needed, metrologists. Every day, the CMMS sends a summary email to our quality assurance
team. If pre-deined conditions are exceeded, the email includes an alert that triggers action by the
team. By plotting the data in a consistent and proven format--the control chart--exceptions and their
associated trends are instantly visible (Figure 1).
Since launching the CMMS, we have observed a number of success stories. The three described here
illustrate the value of the system to us and our customers, and the examples range from "big data"
across multiple sites to a single test at a single site.
Figure 1. This control chart shows actual
calibration data and some of the exceptions
(far right) that triggered rapid action and
resolution.
03 | Keysight | Assurance of Calibration Results: Applying the Power of Visual Information to Improve Quality - Application Note
Scaling the Problem
We have more than 50 service centers around the world. The technicians who work at
those sites calibrate thousands of different models conigured with many possible com-
binations of options. The units they test range from attenuators and digital multimeters
(DMMs) to complex, high-performance microwave signal generators, signal analyzers
and vector network analyzers (Figure 2).
New or old, a typical instrument has 20 to 30 parameters that must be tested. Each of
these includes one to many test points, often resulting in 1,000 measurement points per
calibration for a complex instrument. Whether the calibration occurs in Spain, Singa-
pore or the US, data is collected in a single, structured database. The raw data is stored
electronically and the associated calibration records are retained as an easily accessible
PDF ile 1.
Getting Familiar with the Processes, Tools and People
That mountain of data becomes more meaningful when we can use it to beneit our cus-
tomers. Going from data to information to action requires three things: processes, tools
and people. A closer look at each of these will provide a useful foundation for under-
standing the implementation of our CMMS.
Guiding the process
The concept underlying our process comes from classic Six Sigma thinking: you improve
a process by reducing its variation. By monitoring actual data and lagging exceptions,
we are able to reduce the variations in our processes and ensure valid, reliable calibra-
tion results. To maximize the rigor of our approach, two standards guide us: ISO/IEC
17025 and ANSI/NCSL Z540.3.
ISO/IEC 17025: This international standard deines the general requirements for the
competence needed to perform tests, calibrations, or both. It applies to all organizations
that do such work--irst-, second- or third-party laboratories--regardless of the number
of personnel or the scope of their testing or calibration activities.
The speciication document has ive major sections: scope, normative references, terms
and deinitions, management requirements, and technical requirements. Within the
technical requirements, section 5.9, "Assuring the quality of test and calibration results,"
provides important guidance regarding quality control procedures and the analysis of
quality control data.
Figure 2. Our service centers are equipped
to test virtually every instrument in our
catalog--and they measure every spec, on
every option, every time
1. Our customers can access their calibration records through our free web-based Infoline service.
www.keysight.com/ind/service
04 | Keysight | Assurance of Calibration Results: Applying the Power of Visual Information to Improve Quality - Application Note
Speciically, subsection 5.9.1 deals with monitoring of the validity of calibrations, record-
ing of data to ensure detection of trends, and the techniques used to plan and review the
monitoring activity. Subsection 5.9.2 speciies the need for plans of action in response to
data that falls outside predeined limits. The planned responses must correct the under-
lying problem and prevent the reporting of incorrect results.
ANSI/NCSL Z540.3: This standard deines the requirements for the calibration of test
and measurement equipment. This includes controlling the accuracy of the equipment to
ensure that products and services comply with the requirements.
The Z540.3 handbook has six major sections: scope, references, terms and deinitions,
general requirements, calibration system implementation, and calibration system
assessment and improvement. Within the calibration system requirements, subsection
5.3.2, "Measurement assurance procedures," describes the key attributes of the required
procedures. There are four major themes: thorough documentation of the adopted
procedure; the systematic application of that procedure; criteria that verify attainment
of selected goals; and the need for action, including isolation and correction, when indi-
cated by measurement data. The use of control charts is suggested as a useful tool for
achieving measurement assurance.
Outlining the essential tools
Graphical presentation of data is a long-standing best practice in quality control, in
part because it's an eficient, at-a-glance way to spot trends and otherwise determine
if a process is within statistical control. Well-deined control charts are one of the most
effective ways to present statistical quality control data.
Calibration data typically falls into a normal distribution, often illustrated as the familiar
bell curve (Figure 3).
Figure 3. The normal distribution is the
foundation of the control chart, which
includes the mean,