Technological change is in the air and the buzzwords are as thick as mosquitos on a summer night. Industry 4.0. Machine learning. Artificial intelligence. Digital twins. Analytics. Edge computing. Here are five nuggets of information and suggestions for exploring big data as a foray into digital manufacturing.

1. Big Data Makes for a Good Foundation

In the article, “Digital Transformation on a Small-Business Budget: It Can Be Done,”(1) author Ethan Karp, president & CEO of MAGNET, states that, after watching manufacturers for years, he understands their hesitance to jump on the cutting edge technology bandwagon.

“Most small- and medium-sized businesses believe they’re at a fundamental disadvantage when it comes to Industry 4.0,” Karp said, “largely due to the resources at their disposal. They get overwhelmed, then stuck, putting off tech for a future day.”

Since no one can grasp, absorb and implement all of the new technologies at once, it makes sense to pick a logical starting point that can be built upon later with additional layers of technology. Karp suggests that analytics and big data can be a good first step on a Fourth Industrial Revolution journey. With big data, as with all of the Industry 4.0 advances, said Karp, “You’ll gain a much deeper understanding of the tools and machinery making up your factory floor, with analytics and insights that can be turned into actual business value.”

Implementing big data initiatives offers some rapid results. “When it comes to monitoring real-time data,” Karp said, “the initial return on investment will come quickly. You’ll capture (even) deeper insights as you have more data to pull from, but installing sensors throughout your operation to track and analyze performance will provide immediate upside.”

Sensors give insight into factors such as vibration patterns, pressures, temperature and cooling rates, all of which can be compared to benchmarks or history to see how equipment is performing. That insight can lead to options for finetuning plant operations and launching proactive maintenance procedures. “Start by installing sensors to make your manufacturing operation trackable,” said Karp, “and you may spend just a few thousand dollars while giving yourself access to analytics that will take money off your expense line.”

2. Take Your Pick of Data

Diving into the big data pool is imminently doable. “Manufacturers have data,” said Simon Floyd, industry director, Manufacturing & Transportation, Google Cloud, in Jessie MacAlpine’s article, “What to Do with All That Factory Data? Google Has an Answer.”(2) “The challenge,” Floyd continued, “is contextualization and learning what to take from that data. Companies need to learn how to use data to inform their decision making.”

A single production machine can generate massive amounts of data. Manufacturers can find themselves overwhelmed by the sheer volume of data available, finding it and the cloud computing tools for dealing with it to be much more complex than they expected.

MacAlpine quoted Ford Motor Company’s Jason Ryska, director of Manufacturing Technology Development, who offers clear-eyed advice for avoiding data overload. “First and foremost,” said Ryska, “there is a lot of data that’s available and a [word of] caution in this space [is] that simply because you can collect data on everything, doesn’t mean that you should.” At Ford, Ryska said, they start with a clear problem statement and develop new competencies around that one issue. “Once you get the system and the platform and the tools in place,” said Ryska, “(then) you can scale that to other problems.”

Targeting a straightforward, known problem as a starter project allows manufacturers to focus their attention on a manageable volume of data, to learn the ropes and see the benefits of data analytics.

3. It Takes Skill to Make Sense of Data

Big data can be a big asset, but only if someone has a big understanding of the new trove of data. In his article, “The Importance of Data Literacy And Data Storytelling,” author Bernard Marr cited a data literacy survey which reveals an unrealistic expectation on the part of company leaders.(3) The study was carried out by Forrester Consulting and surveyed 2,000 manager and employers. Based on survey results, wrote Marr, “82% of leaders expect all employees to have basic data literacy, and 79% of leaders say teams are equipping workers with critical data skills – but only 40% of employees say they are being provided with the data skills their employers expect.”

If eight out of 10 company leaders expect data literacy but only four out of ten employees are being trained on that topic, company leaders will never reap the full rewards of big data. Literacy, in this context, includes skill in working with, interpreting, and making decisions based on data. But literacy also extends to knowing how to collect, store and manipulate data, knowing how to assess data’s significance in the small picture and the large picture, and especially how to translate data and its meaning in ways that workers and leaders can understand. Training team members in data literacy will give them the skills to identify valuable data, gather data and convert it as needed, and effectively relay the news – good or bad – that the data shows about, for example, a machine’s condition, the efficiency of a process or the active/idle ratio of a plant’s array of equipment.

4. Choose a Big Data Project Wisely with Smart Input

Ford Motor Company’s Jason Ryska shared more big data initiative advice in Dennis Scimeca’s article, “How Ford Motor Company Handles Big Data.”(4) When making a choice about where to apply big data technology, Ryska suggested turning to a company’s experts for advice on which issues are most amenable to big data solutions. “Don’t let non domain experts or industry define the problem for you,” said Ryska, “because you’re going to get lost in the implementation and data collection. Engage the domain experts and let them define the problem.”

Along with choosing the target area, it also is critical to decide on the scale of the project. While it might be tempting to plan a broad rollout — plant-wide or even across several plants — it might make more sense to pick just one type of equipment or one segment of an operation with which to explore, try out, finetune and implement a big data project. Ryska offers valuable input about starting out modestly. “Start with that domain knowledge and collect realistically what you know with your expertise are the significant variables and parameters,” Ryska said.

It isn’t necessary to collect and analyze all of the available data. Focus on the data that will help solve an identified concrete problem and keep the additional data in mind for addressing other problems later. Ryska suggested choosing the data to include by separating the clearly relevant data from the obviously off-target data and the only potentially useful data. Ignore the off-target data and eliminate the “might be useful” data in favor of the clearly relevant data. More data and variables can be added in later, after analysis of the slam-dunk data.

One last bit of valuable advice from Ryska: get a leg up by learning from others. “If it’s a technology where other companies or other industries are implementing with success, and it’s documented,” said Ryska, “then you should be able to set up clear metrics along the way to guide the team and ensure that you’re achieving similar or better results.”

5. Lean on Big Data

Another approach to tackling a big data project is to use the principles of lean to narrow the scope, to manage the stages of the initiative and to empower a leader to take it from an idea to a pilot project and to ultimate implementation.

In “Drowning in Raw Data? Lean Principles Can Help,”(5) author Torey Penrod-Cambra explored how lean principles can be used to help optimize data collection and prepare it for analysis. “The same lean manufacturing concepts that have transformed manufacturing over the past three decades also apply to data management,” wrote Penrod-Cambra. He focuses closely on purpose, process and people, noting that determining purpose is a great way to start a big data project that will be manageable and successful.

“A lean initiative should target a customer value, such as price, quality or product availability,” Penrod-Cambra wrote. “Information production is no different. Manufacturers should think about who and what their data is serving. Different customers within the organization need access to this information to solve a variety of issues.” Defining the purpose in terms of the customer and the issue to be solved gives a project a tight focus.

Turning to process can facilitate the “how-to” planning of a big data project. “In lean manufacturing,” wrote Penrod-Cambra, “we refer to this as value-stream mapping, which typically involves product and process development, fulfillment and product/customer support.” Process in a big data project entails mapping the flow of data, creating a streamlined and timely data flow, and factoring the “pull” of the project into its design so that data recipients get what they need when they need it.

No amount of purpose identification or process intelligence negates the need for the people – and especially the point person – who will animate the project. Penrod-Cambra explained that a big data project needs a leader who can act as a manager, a champion and an intercompany liaison.

Each of the articles cited here contains more valuable insight and advice for manufacturers who would like to see how big data can improve operations and who want to approach a big data initiative with a plan, confidence and enthusiasm.

References

1. Karp, Ethan. “Digital Transformation on a Small-Business Budget: It Can Be Done,” Industry Week. February 2, 2023. https://www.industryweek.com/technology-and-iiot/article/21259371/industry-40-what-are-manageable-steps-for-smaller-manufacturers
2. MacAlpine, Jessie. “What to Do with All That Factory Data? Google Has an Answer,” Engineering.com. May 29, 2022. https://www.engineering.com/story/what-to-do-with-all-that-factory-data-google-has-an-answer
3. Marr, Bernard. “The Importance Of Data Literacy And Data Storytelling,” Forbes. September 28, 2022. https://www.forbes.com/sites/bernardmarr/2022/09/28/the-importance-of-data-literacy-and-data-storytelling/?ss=enterprisetech&sh=590417ec152f
4. Scimeca, Dennis. “How Ford Motor Company Handles Big Data,” Industry Week. July 14, 2022. https://www.industryweek.com/technology-and-iiot/article/21246494/how-ford-motor-company-handles-big-data
5. Penrod-Cambra, Torey. “Drowning in Raw Data? Lean Principles Can Help,” Industry Week. September 1, 2022. https://www.industryweek.com/technology-and-iiot/article/21249974/drowning-in-raw-data-lean-principles-can-help