0% found this document useful (0 votes)
1K views7 pages

Engineering Statistics

This document discusses statistical techniques used in engineering. It defines statistics and statistical engineering as tools to make quality decisions. The problem solving cycle uses statistics to gather data, form hypotheses, analyze data to make decisions, and apply solutions. Statistical techniques can improve understanding of variability and are powerful decision making tools. Methods to collect engineering data include retrospective studies, observational studies, and designed experiments. Hypothesis testing follows data presentation and is used to make wiser decisions. Analysis uses statistical tests and interpretation to solve problems. Reliability aims to prevent failures through activities like reliability techniques and minimizing maintenance. Maintainability assesses ability to isolate and repair faults using metrics like mean time between failures and mean time to repair.

Uploaded by

hani barakat
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
1K views7 pages

Engineering Statistics

This document discusses statistical techniques used in engineering. It defines statistics and statistical engineering as tools to make quality decisions. The problem solving cycle uses statistics to gather data, form hypotheses, analyze data to make decisions, and apply solutions. Statistical techniques can improve understanding of variability and are powerful decision making tools. Methods to collect engineering data include retrospective studies, observational studies, and designed experiments. Hypothesis testing follows data presentation and is used to make wiser decisions. Analysis uses statistical tests and interpretation to solve problems. Reliability aims to prevent failures through activities like reliability techniques and minimizing maintenance. Maintainability assesses ability to isolate and repair faults using metrics like mean time between failures and mean time to repair.

Uploaded by

hani barakat
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

Literature Review:

Statistics can be defined as the science of making a decision based on collecting


and analyzing numerical. Statistical engineering methods on the other hand are defined
as the tools used by engineers to make decisions in the context of quality management.
The use of statistics in solving problems is evident in the Problem Solving Sycle as this
cycle is comprised of four basic principles:
- Gathering and Collection of Data,
- Presentation and Formulation of hypotheses,
- Analyzing data in order to make wise decisions,
- Applying this decision to solve problems.

Figure 1. Problem Solving Cycle.


It is obvious that having some knowledge of statistics is considered an important
advantage to any engineer. Specifically, statistical techniques which can be used to
design or improve products and systems. These techniques can improve our
understanding of the term variability which indicates a difference in the results of
several observations of one system. Hence, they are considered a powerful tools in the
decision-making processes.

1- Engineering Statistics’ Techniques:


1-1- Collecting Engineering Data
Data can be gathered and collected using three different methods. These
methods are as follows:
1. Retrospective Studies
2. Observational Studies
3. Experiments
1-1-1- Retrospective Study
This methods of collecting data uses existing data recorded over some period of
time. However, this type of study presents some issues:
- It is often hard to establish the relationship between the data collected, because
the rate of these data didn’t change much over the historical period.
- The archived data may not be recorded accurately or there may be a difference
in the methods used to record these data (some data are recorded continuously
and other are recorded hourly). Therefore, we might not be able to construct an
approximate correspondence between them.
Although this method involves large quantities of data, it is found that most of the
time this data offer limited information about the problem. In addition, the margin of
error accompanied with this method is quite big due to the missing of relevant data.
As a result it is hard to establish a reliable explanations of the findings obtained by the
statistical analysis of historical data.
1-1-2- Observational Study
This method requires observing the required data and changing its variables
while recording the resulted changes in the outcome. It is often conducted over short
time periods. Data acquired from observational studies can be useful to accomplish the
following:
1) Provide “real world” information.
2) Detect signals about benefits and risks on respect to the targeted data being
collected.
3) Formulating hypotheses.
1-1-3- Designed Experiments
Designed experiments are considered a powerful approach to studying complex
systems. This method of gathering data requires making deliberate changes in the
variables of the studied system and observe the changes in the outputs in order to
determine which variables are responsible for these changes.
This method is an important tool in the engineering design process because it improves
both the performance and the reliability of the design process while reducing the cost
and the lead time for both the design and development activities.
1-2- Presentation and Formulating a Hypothesis:
After obtaining the data that needs to be studied, we have to turn this row data
into forms and outcomes easy to be used in calculations needed to solve the problem
under investigation. After that a Hypothesis testing is required to help establish some
logical and statistical guidelines in order to make wiser decisions. The goal of the
hypothesis testing is to compare two or more opposing conjectures (hypotheses).
1-2-1- Presentation of Data:
Data are often presented in a way that makes it easy to read and understand by
highlighting their main features using descriptive statistics such as mean and standard
deviation and median and interquartile range.
Another common method to display data is to use Statistical Diagrams and Charts as
they help us interpret these data more efficiently.
1-2-2- Types of Hypotheses:
After presenting the data in an apprehensible form. Formulating Hypotheses is
a vital step in order to make a problem-solving decision. There are several types of
Hypotheses including:
- Research hypothesis: this type represent the outcome the investigator expects to
find.
- Logical hypothesis: this type itself consists of two main categories:
o Null hypothesis: it indicates no significant difference and states that the
observed difference is due to sampling or experimental error.
o Alternative hypothesis: it states that the observed difference is the result
of a real effect.
- Statistical hypothesis: statistical outcome obtained when the research
hypothesis is true.
1-2-3- Basic Steps in any Hypotheses Testing:
1. Defining both null and alternative hypotheses.
2. Selecting the level of significance.
3. Defining decision rules.
4. Compute the appropriate test statistic.
5. Make a decision. State your conclusion in words.

1-3- Analysis of Data:


The next stage is to analyze the data using several types of statistical tests as
follows:
1-3-1- Defining Test Statistic:
In order to check the validity of the hypothesis, one must define the Test
Statistic. This statistic is computed from sample information. After that we calculate
the number that represents the match between a set of data and the expectation outcome
under the null hypothesis.
1-3-2- Probability distribution of the test statistic:
It depends on three parameters:
- The null hypothesis,
- The parameter to be tested,
- The sample size.
Below are the most common types of tests used to define the probability
distribution:
- The Standard Normal Distribution: this distribution represents real-
valued variables with unknown distributions (S.J. Morrison, 2009).
- The Student’s “t” Distribution: this distribution can be used to assess statistical
differences between samples’ means in order to establish confidence
intervals for these differences.
- Chi-square Distribution: This distribution is used to determine the goodness of
fit between an observed and a theoretical distributions (S.J. Morrison, 2009).
- F-distributions.

1-4- Problem Solving:


After formulating hypothesis on the gathered data and analyzing it we can make
a wise decision to solve problems based on the results of these analysis. It should be
noted that statistics need interpretation in order to obtain an answer for the problem.
Another important aspect when solving a problem is to evaluate the validity of the
proposed solution.

2- Reliability in Engineering
As a definition “Reliability” stands for the probability of success which
indicates that the system is performing as designed without any disruptions. Nowadays,
the term “Reliability” is used to describe the probability of certain product or system is
operating successfully for a specified time called (design life) without failure due to
operational and environmental conditions.
From an Engineer’s point of view, “Reliability” can be defined as a collection
of activities interrelating effectively to prevent system function losses. That being said,
Engineering Reliability must concern about eliminating or at least limiting maintenance
requirements, improving both maintenance and productivity using different types of
analysis and problem-solving techniques.

2-1- Important Aspects of Reliability:


1- Applying “Reliability Techniques” such as predictive, analytical methods to
improve the company production and maintenance processes.
2- Minimizing maintenance requirements by implementing PPM programs and
approaches.
3- Advising engineers to consider reliability practices during the design stage to
improve reliability.
4- Updating predictions and test plans during development stage and apply it in
the pre-production stage to make sure all systems are reliable.

3- Maintainability
Maintainability is defined as the probability of isolating a fault in a system and
repairing it by performing maintenance measures to either fixing or replacing the failed
component (s).
Maintainability requires an advanced knowledge in specialized analysis tests
and techniques in order to determine the required time to complete the correct the fault
and restore the system to its original state at the lowest cost (Pham.H, 2006).
Two of the most common indicators used in maintainability studies are:

3-1- Mean Time Between Failures (MTBF):


MTBF refers to the average time elapsed between two consecutive failures (J.
Lienig, H. Bruemmer, 2017). It can be calculated using the formula:
∑(Start of Downtime − Start of Uptime)
MTBF =
Number of Failures

3-2- Mean Time To Repair (MTTR):


MTTR, is the time needed to repair a failure in the system. For any given period
of time (dt), MTTR can be expressed mathematically, as
total maintenance time
MTTR =
number of repairs
MTTR is often used to determine the number of hours that a system will be out
of service due to maintenance process. Hence, it is considered an importance indicator
in engineering reliability as it can provide enough data to make informed decisions such
as repair or replace parts, optimize maintenance schedules, etc.

3-3- Comparison between these two indicators:


These are the main differences between MTBE and MTTR (Stephen Foskett, Pack
Rat, 2016):
- While MTBF measures the reliability of a system, MTTR indicates the
efficiency of the corrective process.
- The higher the value of MTBF the better as it indicates that applied maintenance
methods are being well run. On the contrary MTTR value needs to be as low as
possible since it indicates a high degree of efficiency because it means that the
company is prepared to face problems and present quick solutions to solve them.

3-4- Types of maintenance:


3-4-1- Preventive maintenance (PM):
This type of maintenance refers to the periodic inspection and detection tests in
order to prevent malfunctions before they occur and develop in critical components of
the system.
The main goal of PM is to reduce maintenance costs by minimizing failures in
new equipment using past maintenance data to design a plan with higher reliability,
maintainability (Bamiro, O.A.; Nzediegwu, D.; Oladejo, K.A.; Rahaman, A.; Adebayo,
A. 2011).
3-4-2- Predictive maintenance (PdM)
This type of maintenance concerns in determining the In-Service conditions of
the equipment in order to predict maintenance schedules. This type allow the company
to schedule maintenance processes preventing any unexpected failures (Amruthnath,
Nagdev; Gupta, Tarunm 2018).
3-4-3- Differences between PM and PdM
- In Predictive Maintenance PdM we schedule maintenance tasks based on the
actual condition of equipment. However, in Preventive Maintenance PM the
maintenance processes are scheduled based on the expected life statistics.
- As far as maintenance costs are concerned, preventive maintenance costs more
than predictive maintenance because PM sometimes requires an unnecessary
replacement of a part resulting in more costs while PdM maximize savings that
range through minimizing the stoppage time and cutting down on the price of
spare parts and supplies.

References
- Amruthnath, Nagdev; Gupta, Tarun (2018-02-02). Fault Class Prediction in
Unsupervised Learning using Model-Based Clustering
Approach. doi:10.13140/rg.2.2.22085.14563.
- Bamiro, O.A.; Nzediegwu, D.; Oladejo, K.A.; Rahaman, A.; Adebayo, A.
(2011). Mastery of Technology for Junior School Certificate Examination.
Ibadan: Evans Brothers (Nigeria Publishers) Limited.
- Hoang.Pham (2006). “System Software Reliability”. Springer International
Publishing. pp. 6-7, ISBN 978-1-84628-295-9.
- J. Lienig, H. Bruemmer (2017). "Reliability Analysis". Fundamentals of
Electronic Systems Design. Springer International Publishing. pp. 49–
56. ISBN 978-3-319-55839-4.
- S.J. Morrison (2009), Statistics for Engineers an Introduction, John Wiley &
Sons, Ltd, ISBN: 978-0-470-74556-4.
- Stephen Foskett, Pack Rat. Retrieved 2016-01-18."Defining Failure: What Is
MTTR, MTTF, and MTBF?".

You might also like