Engineering methods, Scientific method, and Data analysis

Action is the foundational key to all success.

Pablo Picasso

To the question “what is an engineer?” the response is usually “an individual that makes automobiles, bridges, medical, equipment, etc.” People has a conception of engineers based on what they produce. They think of engineers in terms of the artifacts they create and see diversity in trades1: A mechanical engineer makes engines, a chemical engineer makes plastics, an industrial engineer improves processes, etc.

Even though these statements are true, they are misleading for our purposes. Our answer to the question is: An engineer is anyone who seeks in her mind, and sets her mental powers in action, to discover or device some means of succeeding in a difficult task she might have to perform.

It is less about the artifacts and more about how engineers go about creating or improving artifacts, processes and systems. The different ways in which we can create or improve are called engineering methods or processes.

An engineering method can be defined broadly as a process to create the best change in a poorly understood or uncertain situation under the constraint of available resources.

An engineering method applies heuristics2. A heuristic can be understood as anything that provides a plausible, usually not optimal but acceptable, aid or direction in the solution of a problem.

Engineering heuristics are continuously evolving as engineers and scientists create and develop new techniques. The following sections will briefly describe some of the more important engineering methods and heuristics.

Engineering method: Toyota Improvement Kata

This method was developed and refined by Toyota [@rother2010] and is considered a strategic approach to problem solving, improvement, design and innovation. It starts with two elements:

After we state where we are and where we want to be: the gap, we can use engineering tools and processes to close it.

An Engineering Kata can be illustrated with four steps:

  1. Understanding the direction to go or challenge to overcome. This will set the vision of what is required to accomplish, and it is usually declared in non quantitative and aspirational terms.

  2. Quantifying the current condition. This step can also be seen as the quantification of a baseline or start position for our journey.

  3. Establishing a quantifiable target condition that will focus on measurable results rather than random improvements. This target condition is an intermediate goal between the baseline and the challenge. The development of target conditions creates a series of gaps that are easy to solve: a natural partition of our challenge in manageable bites.

  4. Conduct engineering experiments to achieve the target condition. In this step we use engineering tools and methods to close the gaps.

  5. Repeat the steps until the challenge is overcome.

This process is presented in @fig-engKata.

Toyota Improvement Kata and feedback to stakeholders.
Toyota Improvement Kata and feedback to stakeholders.

The Toyota Improvement Kata helps develop fundamental skills of working like a scientist in an engineering environment. The learner iterates or experiments his or her way toward a desired goal instead of simply deciding the way forward. This way of thinking and working helps people successfully deal with uncertainty and challenges.3

Apollo: A challenge and a program

The challenge4 is usually expressed in layman terms and not well specified from an engineering perspective: we do not know how to overcome the challenge…yet.

As an example, consider the memorable objective set by John F. Kennedy in 1961 during his appearance at the joint session of congress.

{{< video https://www.youtube.com/watch?v=GmN1wO_24Ao&t=125s title=‘The Goal to go to the moon?’ start=‘100’ aspect-ratio=“21x9” >}}

JFK: Excerpt to joint session of congress, May 25th, 1961

…First. I believe that this nation should commit itself to achieving the goal, before this decade is out, of landing a man on the moon and returning him safely to the earth. No single space project in this period will be more impressive to mankind, or more important for the long-range exploration of space; and none will be so difficult or expensive to accomplish. Ref: jfklibrary.org

@fig-Apollo presents a summary of the Apollo program. The participation of engineers and engineering methods was crucial to overcome the challenge.

Apollo program: progression to final objective.
Apollo program: progression to final objective.

Engineering method: Loyola’s five elements

This method was developed by Ignatius of Loyola, who was a Spanish Catholic priest and theologian canonized saint in 1622. As a former soldier, Ignatius paid particular attention to the spiritual formation of his recruits and recorded his method in “Spiritual Exercises” [@loyola1879]. The main objective of the exercises was to influence spirituality and discernment, we will briefly comment on the elements of his method to achieve discernment5.

Loyola’s method consists of five elements:

  1. Observe,

  2. Think,

  3. Act,

  4. Evaluate, and

  5. Context.

These elements interact with each other and are represented in @fig-ignatius:

Ignatius Method: Context defines the tools and techniques used for observation, thinking, acting, and evaluating.
Ignatius Method: Context defines the tools and techniques used for observation, thinking, acting, and evaluating.

If an engineer wants to close a gap6, the first step is to observe the process or system under investigation. After and during the observation, the engineer thinks of ways to solve the gap and develops a plan of action. The engineer will evaluate his actions and proceed to adjust the plan accordingly. She will also acquire knowledge that will be useful to solve future problems. The arrows in @fig-ignatius show that there is no specific sequence of activities to follow. Nevertheless, all elements need to be covered to successfully solve a problem.

Context, the fifth element determines the tools to be used when going through the other four elements. For example, observation in biology might imply the use of microscopes, while in subtractive manufacturing we will need to use gauges.

Engineering method: PDCA

The PDCA cycle (Plan, Do, Check, and Act) was developed by Walter Shewhart an popularized by W. Edwards Deming [@shewhart1931]. These four phases are:

Plan: Here the engineers define the problem or gap. They should identify the desired outcome, and develop a plan to achieve it.

Do: Where engineers implement the plan. They collect data on the process and make sure that the plan is working as expected.

Check: The engineers evaluate the results of the Do step comparing their results to the desired outcome, and identifying any areas that are still in need of improvement.

Act: Engineers make changes to the plan based on the results of the Check step. They implement the changes and repeat the cycle until the desired outcome is achieved.

The PDCA cycle is a simple but powerful tool that is used to improve any process. This method is also known as the Deming Cycle.

PDCA cycle is a valuable process that can be used by organizations and individuals to improve processes and achieve goals. It is a simple but powerful tool that can be used to make a real difference, @fig-PDCA.

PDCA or Deming Cycle: Plan, Do, Check, Act.
PDCA or Deming Cycle: Plan, Do, Check, Act.

An example using PDCA is presented in the following video from Skanska group:

{{< video https://www.youtube.com/watch?v=qHVxNH0v7wY title=‘Safety approach:PDCA walk’ aspect-ratio=“21x9” >}}

Engineering method: DMAIC

The DMAIC methodology (Define, Measure, Analyze, Improve, and Control) was not invented by a single person. It is a combination of different quality improvement methods that were developed over time. However, the two people who are most commonly credited with the development of DMAIC are Mikel Harry and Bill Smith[@harry2005].

Harry and Smith were working at Motorola in the 1980s when they developed the MAIC methodology, which was the predecessor to DMAIC. MAIC stands for Measure, Analyze, Improve, and Control. The DMAIC methodology was later developed by adding a Define step at the beginning of the process.

DMAIC is a data-driven approach to improving business processes sequentially:

Define: In this step the team defines the problem or gap. They identify the customer needs, and define the current state of the process.

Measure: The team performs two tasks:

  1. Validate the measurement devices used to evaluate the performance of the process, and

  2. Measure the current state of the process collecting data on inputs, outputs, and performance metrics.

Analyze: The team uses statistical tools to analyze the data and identify the root causes of the problem. The objective is to identify two types of relations:

\[ y_{operational}=f(\textbf{x}) \]

and

\[ y_{financial}=f(y_{operational}) \]

@eq-opery is a measure of the process’s operational performance under analysis and is a function of \(\textbf{x}\): the inputs of the process.7 @eq-yfin is the transformation of this performance in terms of expected financial impact.

Improve: The team implements changes to the process to address the root causes of the problem. They test the changes and make sure that they are effective.

Control: The team puts in place controls to ensure that the performance and financial impact of the improvements are sustained. The team monitors the process for a period of time and make adjustments as needed. An example of DMAIC is presented in the next video:

{{< video https://www.youtube.com/watch?v=S72T2ThAYyM title=‘DMAIC’ aspect-ratio=“21x9” >}}

Engineering method: Design for Six Sigma (DFSS)

Design for Six Sigma (DFSS) is a systematic approach to designing new products and services that meet or exceed customer requirements, while also being manufacturable and cost-effective . It is a variation of the Six Sigma methodology, and is used to prevent defects and improve quality during the early stages of product development.

DFSS is based on the following principles:

Customer focus: The goal is to create products and services that meet or exceed customer needs.

Design for manufacturability8: The products and services created using DFSS are easy to manufacture and assemble.

Design for cost-effectiveness: The products and services designed are cost-effective during production and delivery.

DFSS is deployed following a five-step process called DMADV:

Define: Clearly specify and characterize all vital customer requirements and the processes, products or services to meet those requirements.

Measure: Identify and quantitatively estimate the gaps between the current state and the desired state of the process, product or service.

Analyze: Identify the root causes of the gaps and develop solutions to close them.

Design: Create a new product or service that meets the customer requirements.

Verify: Confirm that the new product or service meets the customer requirements, is manufacturable, and cost-effective.

DFSS is a complex process that requires a team of experts with different skills and knowledge. It takes time to implement and can be expensive.

DFSS is not a guarantee of success but it improves the chances of success. A summary of DMADV is presented in the next video:

{{< video https://www.youtube.com/watch?v=rcZoQkjw2js title=‘DMADV’ aspect-ratio=“21x9” >}}

Engineering method: TBP

The Toyota Business Process Method (TBP) is a systematic approach to improving business processes. It is based on the Toyota Production System (TPS), which is a set of principles and practices that Toyota uses to manufacture its products. TBP is a five-step process:

  1. Clarify the gap: The first step is to clarify the problem to solve. What is the current state of the process? What are the desired results?

  2. Break down the problem: Divide the problem into smaller, more manageable pieces. This will make easier to understand understand the root causes of the problem and to develop solutions.

  3. Set a target: A quantitative target answers the question: what are we trying to achieve by improving the process?

  4. Analyze the root causes: Answer the question: what are the factors that are contributing to the problem?

  5. Develop countermeasures9: These countermeasures should be specific, measurable, achievable, relevant, and time-bound.

  6. See countermeasures through. Be sure that all countermeasures are put in place as planned.

  7. Evaluate and monitor the results and processes. Verify that all measures and processes are consistently within the expected range and performance.

  8. Standardize success. Document and communicate lessons learnt.

@fig-relation presents the relation between TBP, DMAIC and PDCA.

PDCA, DMAIC, and TBP relation.
PDCA, DMAIC, and TBP relation.

The Scientific Method

Engineers also use science’s Queen process: The Scientific Method.

This is a process for acquiring knowledge and explain nature that has characterized the development of science since at least the 17th century10. It involves careful observation, and application of rigorous skepticism about what is observed, given that our cognitive assumptions can distort how one interprets the observation. A video of Dr. Jim Allison11 provides an interesting explanation of the scientific method:

{{< video https://www.youtube.com/watch?v=Myjz7PzgybY title=‘Scientific method by Jim Allison’ aspect-ratio=“21x9” >}}

The scientific method foundation is the gathering of empirical evidence from experiments and observations, and using this evidence to form hypotheses and theories.

The scientific method is often described as a series of steps, but it is important to remember that the process is not always linear. Scientists may go back and forth between steps, or they may need to repeat the entire process if they find that their results are not what they expected.

The following video explains the scientific method with an example:

{{< video https://www.youtube.com/watch?v=Xxm_beTs2LU&t=43s title=‘Scientific method example’ aspect-ratio=“21x9” >}}

Engineering design and problem solving method: EDPS

The engineering design and problem solving approach (EDPS) is a systematic process that can be used in a variety of fields, including engineering, science, and business.

The EDPS method typically consists of the following steps:

Define the problem: The first step is to define the problem that you are trying to solve. This includes understanding the root cause of the problem and the desired outcome.

Gather information, research: The next step is to gather information about the problem. This includes collecting data, researching the problem, and talking to experts.

Generate possible solutions: Once information is obtained, the engineer generates solutions to the problem. This includes ideation, sketching, and prototyping.

Evaluate solutions and select candidates for implementation: This includes testing the solutions, analyzing the results, and making improvements.

Implement the solution: The final step is to implement the solution chosen. This includes implementing final changes and adaptations to the prototype, testing these changes on the field, and monitoring the results.

EDPS is iterative, meaning that it is mostly a process of educated trial and error that involves planned experimentation and tinkering. The following video presents a variation of the engineering process:

{{< video https://www.youtube.com/watch?v=fxJWin195kU title=‘Engineering method example’ aspect-ratio=“21x9” >}}

Remember: we engineers learn from failure, we keep going.

Data Science and Engineering

Engineering is one industry that has been particularly influenced by the growing need for data collection and analysis. As big data has begun to play a larger role in industries around the world, engineers have been called on to play an influential role in the way this information is gathered, stored, and leveraged. Professionals with an engineering background generally prove to be particularly adept at developing techniques for analyzing data groups to extract valuable information.

To succeed in a career as a data scientist, an engineer should possess the following qualifications:

The engineering method

Engineering Data Analysis

Engineering Data Analysis (EDA) is a field that includes concepts, tools and techniques used to extract insights and information from data to achieve a desired outcome. These insights and information are used to act on products, services, processes, and systems.

Data analysis is not as straightforward as applying formulas and obtaining numbers. It requires, as a minimum, a deep understanding of the data and the process under analysis, honed ability to think critically and creatively, statistical and subject matter skills, as well as the ability to communicate our findings: data storytelling. With the state of engineering and science today, it is very difficult to be proficient in all the areas mentioned above: that is the reason for working in teams.

EDA is used to engineer things. EDA is part of all the engineering and scientific methods mentioned in this chapter. An engineer worth his or her weight in salt is a good data analyst.

  1. This section uses concepts and ideas on the engineering method developed by [@koen2009].↩︎

  2. The term heuristic originates from the Greek verb heurisko, which means to discover or find out.↩︎

  3. Reference: A very good summary can be found in: https://rb.gy/5j9vp↩︎

  4. Generally a strategic goal.↩︎

  5. Discernment is the ability to perceive, understand, and judge things clearly, especially those that are not obvious or straightforward.↩︎

  6. It is practically tractable to define a problem as a gap to be closed. A gap being the separation between the actual and desired state.↩︎

  7. These inputs can be controllable or uncontrollable (noise).↩︎

  8. The degree to which a product can be effectively manufactured given its design, cost, and distribution requirements.↩︎

  9. At Toyota the term countermeasure is used to describe a proposed solution. Toyota’s belief is that problems are never solved completely and countermeasures only mitigate the effects of the problem analyzed.↩︎

  10. In the West. For sure, other regions of the world have applied this method in one fashion or another.↩︎

  11. In 2018, Dr. Jim Allison was awarded the Nobel Prize in medicine for discovering an effective way to attack cancer through immunology. In his lab, Allison urges researchers to get rid of the idea that they can prove something with science. All they can do is fail to disprove.↩︎