Lean Digital Thread: Data-Driven Decisions and Micro-Solutions in Manufacturing

Believe it or not, I’m going to quote Robin S. Sharma in this column: “Small daily improvements over time lead to stunning results.” It would be even easier to make a point with compounded interest, and you all know about Kaizen (the Japanese business concept focused on continuous improvement across the organization from the CEO down to the shop floor operators).

People and factories have been collecting data, verifying it, and translating it into reports for a long time, but context always makes a difference. People will sometimes interpret and challenge the outcome; other times, they will try to validate and verify the accuracy of the data (smart and required, but it’s a slippery slope). However, some will create additional reports, detect the root cause of the problem, remove the outliers, and improve, or even use the data to innovate.

In past columns, I’ve written about two pillars:

  1. Data collection and the basic questions you can answer
  2. Material management and its impact

In this column, I’ll discuss the next level—changing the mindset from reporting to analytics and focusing on making small improvements.

Data-Driven Decision Making?
We all have the importance of data-driven decision making in mind, but studies show that most companies still make decisions based on gut feelings or previous experiences (Figure 1).

sagi_fig1_infographic.jpg

Figure 1: Infographic.

In my opinion, it’s partially related to ego (but not always in a bad context because self-esteem is also based on past experience) and confidence, which is fine since it’s an important part of our personality.

Here’s another quote by the one and only Dan Ariely, professor of psychology and behavioral economics at Duke University: “individuals are honest only to the extent that suits them (including their desire to please others).” According to studies in neuroscience, we consistently attend to irrelevant information too much and even rationalize our bad decisions. Many of those biases—rational or irrational—operate outside of our awareness pretty quickly. Merely recognizing that we have a bias does not always make that bias go away.

Let’s go back to manufacturing. Whatever data collection system is used, any effort to digitalize will need to engage and empower the production team at the factory; that’s clear. One must attend to the manufacturing process but also act as the front line of communications and control (as many factories still use Excel, this aspect is crucial).

We all want the same thing—to improve the way we do things; it’s as simple as that. Thus, there are two scenarios:

  1. Reactive: Something happened; in this case, we need to minimize the effect of the problem on the final product or process through problem-solving and corrective actions in real-time. In some cases, we need to provide intelligence when engineering is called to the work cell so that they can spend less of their time on data gathering and more focus spent on re-engineering for root-cause elimination to remove the possibility for any undesired condition to reappear.
  2. Proactive: Nothing happened, at least according to the current KPIs. What do you do next? In manufacturing (or real life), the answer is too broad. Narrowing down the options is an important step, and based on the feedback I collected, without narrowing down the options, it’s almost impossible to improve.

TMI (Too Much Information)
I’ll start with an anecdote. As one of our customers started using our manufacturing analytics platform, one of the data scientists met them and tried to get some more insights so that we could develop new “dossiers.” They came up with a list of 20 reports, claiming that they are all super important. We provided some of the reports, but we measured how many times they were opened (I’m not even talking about being used); as we anticipated, it wasn’t much. As in real life, we are often overwhelmed with data from multiple sources (Figure 2).

Sagi_fig2_evolution.jpg

Figure 2: How can we deal with data overload?

How can we separate the wheat from the chaff? You need to invest in an analytics platform and data collection—that’s clear—but very few factories can hire a dedicated domain expert and/or data scientist to dig in the data 24/7 to extract advanced insights. We may have another path in the digitalization journey with some kind of a results-oriented shortcut through the ISA-95 levels.

Micro-Solutions
Some of you may have heard the term micro-services—a rising software architecture style. The concept is to develop small applications that will focus on narrow functionalities. I’m suggesting you take the same approach to the manufacturing floor. There’s still a huge need for monolithic platforms, but we can also improve some KPIs with small applications that will run at the shop floor level. (As quoted earlier, “Small daily improvements over time lead to stunning results.”) The idea is to answer one question at a time, and the outcome should be a slight improvement to the yield.

Summary
So far, we’ve discussed:

  1. Data collection and the basic questions you can answer
  2. Material management and its impact
  3. Data-driven decision making vs. decisions based on a gut feeling
  4. An introduction to micro-solutions

Next Time
Stay tuned! In my next column, I will present a few micro-solutions for electronics manufacturing. And feel free to drop me a note on LinkedIn if you have an interesting topic to discuss.

References

Sagi Reuven is a business development manager for the electronics industry, Siemens Digital Industries. Download your free copy of the book The Printed Circuit Assembler's Guide to… Advanced Manufacturing in the Digital Age from Mentor, a Siemens Business, and visit I-007eBooks.com for other free, educational titles.

Back

2020

Lean Digital Thread: Data-Driven Decisions and Micro-Solutions in Manufacturing

08-26-2020

In past columns, Sagi Reuven has written about two pillars: (1) data collection and the basic questions you can answer, and (2) material management and its impact. In this column, he discusses the next level—changing the mindset from reporting to analytics and focusing on making small improvements.

View Story

Lean Digital Thread: Driving Productivity Excellence—Lean PCB Manufacturing

07-22-2020

During the last few months, thanks to COVID-19, Sagi Reuven felt the supply chain impact on our hardware production for the first time and hence dedicates this column to effective material management.

View Story

Lean Digital Thread: Realizing ROI Starts With Making Smart Data

06-16-2020

Sometimes, you think too much about the bits and bytes, and you forget why you started this journey; we’re all caught up in a tornado of marketing buzzwords. In Sagi Reuven's debut column, he shares how initialization has to start by turning data into smart data.

View Story
Copyright © 2020 I-Connect007. All rights reserved.