Tema 1 Introduction To Measurement Theory-S2020 - First Part

ETSETB-DEE-UPC ELECTRONIC INSTRUMENTATION & OPTOELECTRONICS
Unit 1. Introduction to measurement theory

• 1.1 Basic metrological terms and concepts
• 1.1.1 What is to measure?
• 1.1.2 Some definitions
• 1.1.3 Source and error classifications
• 1.2 Methods for evaluation of measurement uncertainty

• 2.2.1 Introduction
• 2.2.2 Mathematical modeling of uncertainty
• 2.2.3 Type A uncertainty evaluation
• 2.2.4 Type A uncertainty evaluation
• 2.2.5 Combined uncertainty
• 2.2.6 Uncertainty Budget
1 Juan José Ramos Castro & Mireya Fernández Chimeno

Fiona Auty, Head of Government Relations and Corporate Communications at NPL
Measurement is at the heart of all science and engineering. It is only when we can measure
something that scientists can study it and engineers can improve it. And since science and
engineering play an important role in our lives, measurement matters for everyone.
Measurement affects our daily lives:
•When we buy a part that ‘just fits’: a nut fits a bolt, or a Lego®
brick sticks perfectly to another brick.
•When our medical care depends critically on measurements of

concentrations of chemicals in blood, or the intensity of X-rays
•When a satellite navigation system guides us along a road, and it

depends on time measured by ultra-precision clocks on satellites
In all these situations, and thousands more, we are enjoying the benefits of a global
system of measurement.

1.1 What is to measure?
L1 ? 333 ֍
222 ֍
L2 ?
La ? Number Measurement unit
999 ☺
666
☺
Lb ?
La ?
999 ☺
Measurement Procedure
Lb ?
MEASURE: assign objectively and empirically a

number or symbol to a quality/attribute of an object or
event in a representative way.

Numerical assignment:
comparison of the quality to
be measured with a
reference standard.
Lb = 18,5 squares

Juan José Ramos Castro & Mireya Fernández Chimeno

During the 26th meeting of the General Conference on

Weights and Measures (CGPM), the Member States of the
Metre Convention voted to revise the International System of
Units (SI).
This means that the base units are now defined in terms of
constants that describe the natural world, which are the most
stable references available for us to use.
From https://www.npl.co.uk/si-units/the-redefinition-of-the-si-units


Examples of incorrect use of units:
Crash of Mars Climate Orbiter Observer spacecraft:
Robotic space probe launched by NASA on December 11, 1998 to study the Martian climate.
Programming teams in Europe and USA use two measurement systems, imperial and metric, to
calculate the trajectory of the spacecraft. The probe entered in the Martian atmosphere at the
wrong angle, and promptly disintegrated.


Examples of incorrect use of units:
The ‘Gimli Glider’.

An Air Canada Boeing 767-233 jet was refueled in Montreal using 22 300 pounds of fuel instead of
22 300 kilograms. The pilot calculated how much fuel he needed thinking he was getting his fuel in
pounds per liter. When the plane ran out of fuel mid-flight, the pilot had to make an emergency
'gliding' landing at Gimli Canadian Air Force Base.
Why?
At the time of the incident, Canada's aviation sector was in the process of converting to the metric
system. As part of this process, the new 767s being acquired by Air Canada were the first to be
calibrated for metric units (litres and kilograms) instead of Imperial units (gallons and pounds). All
other aircraft of the company were still operating with Imperial units.
See the document “the Gimli Glider” in Atenea for more information

1.2 Some definitions
Measurement is only useful if it is suitable for the intended purpose.
Consider the following questions:

• Do you know how accurate your measurement result is?
• Is this accurate enough?
• How strongly do you trust the result?
These questions relate to the quality of a measurement. When talking

about measurement quality, it is important to understand the following
concepts:
Precision, accuracy, uncertainty,
repeatability, reproducibility,
tolerance, traceability and calibration
Traceability and Calibration

Traceability and Calibration

Measurements are only traceable if the calibration corrections and
uncertainties are actually applied at every step in the chain back to the
reference.

Accuracy, Precision and Uncertainty
Precision is about how close measurements are to one another. Accuracy is

about how close measurements are to the ‘true value’.
In reality, it is not possible to know the ‘true value’ and so we introduce the
concept of uncertainty to help quantify how wrong our value might be

The difference between accuracy and precision is illustrated here.
Accuracy is a qualitative measure of how close a measurement is to the

center of the target – the ‘true value’. Precision is represented by a cluster
of consistent measurements, but there is no guarantee that these are
accurate.

In practice we are not able to view the target and assess how close to the
‘true value’ our measurements are.
What interests us is the answer to the question “How far from the target
could our arrows have fallen?” We also need to ask “How wrong could
we have been?”
To answer these questions we need to look at all the factors that go into
making a measurement and how each factor could have affected the
final estimate of the answer.
The answer to “How wrong are we likely to have been?” is known as the
‘measurement uncertainty’, and this is the most useful assessment of
how far our estimate is likely to lie from the ‘true value’.

Don't confuse mistakes with errors!

Measurement scientists use the term ‘error’ to specify the difference between an
estimate of quantity and its ‘true value’. The word 'error' does not imply that any
mistakes have been made.
Where the size and effect of an error are known (e.g. from a calibration certificate) a
correction can be applied to the measurement result. If the value of an error is not
known, this is a source of uncertainty.

Repeatability and Reproducibility
If you take a voltage measurement three times in one minute using the same multimeter, you
would expect to get a similar answer each time. Repeatability describes the agreement within
sets of measurements where the same person uses the same equipment in the same way,
under the same conditions.
But if your colleagues each had a go at taking the same measurement on different days using
different measuring equipment, a wider range of answers would be much less surprising. This
is known as 'reproducibility' and describes the agreement within a set of measurements
where different people, equipment, methods, locations or conditions are involved.

Tolerance
How is it that nuts from one supplier and bolts from another supplier work
together? The answer lies in tolerance, also known as 'acceptance criteria'.
The tolerance is the agreed allowable variation in the shape of the nuts and
bolts that allow them to still fit together.
Tolerance is the maximum acceptable difference between the actual value

of a quantity and the value specified for it.
For example, if an electrical resistor has a specification of 10 ohms and

there is a tolerance of ±10 % on that specification, the minimum acceptable
resistance would be 9 ohms and the maximum would be 11 ohms.

Direct measurement: that obtained from a single reading of a

measuring instrument
 R = 100 
Indirect Measurement. that obtained from various direct measurements

and a law that relates them.
+
V I
1V
V V 1V
-
R   100Ω
I 10 mA
10 mA
I
MEASUREMENT BY COMPARISON. The

measured variable is compared with a reference
quantity of the same type
Ex precision scale
MEASUREMENT BY DEFLECTION. is obtained

from various direct measures and a law that relates
Ex Bathroom Scale
[N] Load [m] Strain [] Wheatstone [V] [V] Weight [g]
gravity Amplifier A/D
M Cell Gauge Bridge
Vref [V]
Juan José Ramos Castro & Mireya Fernández Chimeno

1.3 Source and error classifications
What affects your measurements?
Many factors can reduce accuracy or precision and increase the

uncertainty of your measurement result:
• Environmental conditions – changes in temperature or humidity can

expand or contract materials as well as affect the performance of
measurement equipment.
• Inferior measuring equipment – equipment which is poorly
maintained, damaged or not calibrated will give less reliable results.
• Poor measuring techniques – having consistent procedures for your
measurements is vital.
• Inadequate staff training – not knowing how to make the right
measurement, not having the confidence to challenge the results and
not being willing to seek advice can all have a negative impact.

Types of errors
Two types of effects can modify the value of the readings of the
instruments: the systematic effects and random effects. Also, occasionally
absurd readings (aberrant readings may appear).
Aberrant
Random reading
effects
Systematic
effects

Bias
After removal the aberrant readings, one qi reading of one magnitude
q can be modeled generally as:
qi=q+b(q)+c(q,i)
b(q) is the systematic effect and c(q,i) the random effect.
b(q): bias
- Constant (if time for measurement is short)

- Generated by drifts in the instrument due to factors that can
not be controlled (aging, temperature, ...), or inherent to the
method of measurement (loading effect).
- If your presence is known they must be corrected before
attempting to analyze the uncertainty in the measurement.

Random effects
If you have rejected the aberrant readings and corrected the systematic
effects the only difference between the readings and the value of the
quantity that you want measure is due to random effects.
-Characterized from its probability distribution function, which is

usually symmetric
-Values are bounded
-Its mathematical expectation is zero
By averaging several readings the random effects are reduced if the

readings are statistically independent. (averaging theoretically an infinite
number of readings random effects are completely eliminated).

Example of systematic and random effects on Gravitational constant

measurement in the last 100 years.
NIST recommended value (2014)

G=6.67408±0.00031×10−11 Nꞏkg–2ꞏm2 (relative standard uncertainty 46 ppm)
Systematic
effect with
respect
recommended
value

Tema 1 Introduction To Measurement Theory-S2020 - First Part

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tema 1 Introduction To Measurement Theory-S2020 - First Part

Uploaded by

Copyright:

Available Formats

ETSETB-DEE-UPC ELECTRONIC INSTRUMENTATION & OPTOELECTRONICS

Unit 1. Introduction to measurement theory

• 1.2 Methods for evaluation of measurement uncertainty

1 Juan José Ramos Castro & Mireya Fernández Chimeno

Fiona Auty, Head of Government Relations and Corporate Communications at NPL

•When our medical care depends critically on measurements of

•When a satellite navigation system guides us along a road, and it

2 Juan José Ramos Castro & Mireya Fernández Chimeno

1.1 What is to measure?

1.1 What is to measure?

MEASURE: assign objectively and empirically a

4 Juan José Ramos Castro & Mireya Fernández Chimeno

1.1 What is to measure?

5 Juan José Ramos Castro & Mireya Fernández Chimeno

1.1 What is to measure?

Juan José Ramos Castro & Mireya Fernández Chimeno

During the 26th meeting of the General Conference on

7 Juan José Ramos Castro & Mireya Fernández Chimeno

1.1 What is to measure?

Crash of Mars Climate Orbiter Observer spacecraft:

8 Juan José Ramos Castro & Mireya Fernández Chimeno

1.1 What is to measure?

The ‘Gimli Glider’.

9 Juan José Ramos Castro & Mireya Fernández Chimeno

1.2 Some definitions

Measurement is only useful if it is suitable for the intended purpose.

Consider the following questions:

These questions relate to the quality of a measurement. When talking

Traceability and Calibration

11 Juan José Ramos Castro & Mireya Fernández Chimeno

Traceability and Calibration

12 Juan José Ramos Castro & Mireya Fernández Chimeno

Accuracy, Precision and Uncertainty

Precision is about how close measurements are to one another. Accuracy is

13 Juan José Ramos Castro & Mireya Fernández Chimeno

Accuracy, Precision and Uncertainty

The difference between accuracy and precision is illustrated here.

Accuracy is a qualitative measure of how close a measurement is to the

14 Juan José Ramos Castro & Mireya Fernández Chimeno

Accuracy, Precision and Uncertainty

15 Juan José Ramos Castro & Mireya Fernández Chimeno

Don't confuse mistakes with errors!

16 Juan José Ramos Castro & Mireya Fernández Chimeno

Repeatability and Reproducibility

17 Juan José Ramos Castro & Mireya Fernández Chimeno

Tolerance is the maximum acceptable difference between the actual value

For example, if an electrical resistor has a specification of 10 ohms and

18 Juan José Ramos Castro & Mireya Fernández Chimeno

Direct measurement: that obtained from a single reading of a

Indirect Measurement. that obtained from various direct measurements

MEASUREMENT BY COMPARISON. The

MEASUREMENT BY DEFLECTION. is obtained

Juan José Ramos Castro & Mireya Fernández Chimeno

1.3 Source and error classifications

What affects your measurements?

Many factors can reduce accuracy or precision and increase the

• Environmental conditions – changes in temperature or humidity can

21 Juan José Ramos Castro & Mireya Fernández Chimeno

22 Juan José Ramos Castro & Mireya Fernández Chimeno

b(q) is the systematic effect and c(q,i) the random effect.

- Constant (if time for measurement is short)

23 Juan José Ramos Castro & Mireya Fernández Chimeno

-Characterized from its probability distribution function, which is