Big data meets manufacturing: Beyond the buzzwords

Big data in manufacturing - electronics

There’s an old-school saying still popular today. “Too many cooks spoil the broth.” In the workplace, this translates to a familiar scene: where there are too many people giving instructions and often not enough people doing the work.

Today, big data can be equated with some of those cooks and in this case, too many of them. Big data has taken over and redefined the manufacturing industry. It’s moved beyond the buzzword to become a highly effective growth tool, one that informs our industry’s decision-makers, helping them to make better decisions and drive performance.

Define it (and tell me what it is). Big data is the large volume of data, structured and unstructured, that inundates a business on a day-to-day basis. Big data in manufacturing or rather, big data analytics in manufacturing is about more than the amount of data in the industry. We like the way Matthew Littlefield at LNS Research, says it:

“Big data in manufacturing is about using a common data model to combine structured business system data with structured operational system data with unstructured internal and external data to uncover new insights through advanced analytical tools.”

Implement it (in manufacturing). For starters, it’s not the amount of data that’s important, it’s what organizations do with the data that matters. Industrial, automotive and hi-tech manufacturers use big data to design and test new products, optimize R&D, improve processes, reduce product defects and optimize their supply chains, marketing, customer service and finance.

In a Forbes article, although not new, industry analyst firm, Gartner, makes big even bigger with words like volume (the amount of data), velocity (the speed of information generated and flowing into the enterprise) and variety (the kind of data available). Gartner also proclaims, “Every enterprise needs to fully understand big data – what it is to them, what is does for them, what it means to them –and the potential of data-driven marketing.”

Challenge it (and make it accurate). There are numerous reasons why big data is hard to get our heads around and although we’re all in this together, it doesn’t make it any easier.

For one thing, big data must be super-fast. Not only must we find and analyze the relevant data we need, we must find it quickly. Then we have to understand the data and get it into top-notch shape so we can analyze it. But even if we manage to find it and analyze it and put it in the ‘right’ context, it’s not worth much if it’s not accurate.

Why big is ‘big’ with manufacturers

Big data helps predict future events. This alone, helps manufacturers improve margins, boost energy efficiency and sustainability, address regulatory concerns, increase product quality and reliability, save on costly IT and human resources.

The first step for manufacturers that want to use big data analytics is to see how much data the company has at its disposal. Most companies collect volumes of process data but typically use it only for tracking purposes, not as a basis for improving operations. For these players, the challenge is to invest in the systems and skillsets that will allow them to optimize their existing process information. For example, by centralizing data from multiple sources, data analysts can draw actionable insights and useful, decision-making information.

According to McKinsey, “But it’s still only a tool. What matters is just how effective big data can be when it’s used right.” Well, since there’s no shortage of big data in manufacturing success stories, we’ve cherry-picked a few to bring our point home or rather – to the shop floor.

Maximized processes

Key analysts at McKinsey often cite big data success in pharma manufacturing, in this case, a biopharma company. The company deployed live, genetically-engineered cells, tracking 200 variables to assess the purity of its manufacturing process for vaccines and blood components. However, two batches of the same substance manufactured using identical processes showed a yield variation of 50-100%. The project team divided its manufacturing processes into clusters and using big data analytics, identified 9 parameters that had a direct impact on vaccine yield. By modifying target processes, they increased vaccine production by 50% – saving between $5-10 million annually.

Enhanced QA

Intel must test every chip that rolls off its assembly lines, running each one through a battery of 19,000 tests. Recently, by implementing big data in manufacturing process, the QA team yielded actionable insights, enabling them to cut test time and reduce the number of tests required. Implementing big data analytics saved $3 million in manufacturing costs for a single line of Intel Core processors. By expanding big data use in its chip manufacturing, Intel expected to save an additional $30 million.

Boarding the big data train

Big data isn’t all technology – it’s also strategy and if you’ve got both, then you’re off to a fine start. To give big data a ‘big welcome’ in your organization, here are some true n’ tried steps to help you towards your goal:

  1. Define a strategy – Set your priorities and goals and build a roadmap – define what you’re looking to achieve
  2. All aboard! – Get the right players in your court, all C-level management, IT and an outside consultant if needed
  3. Set / define your CSFs – Critical Success Factors determine the value of your project from the get-go, so establish, implement and agree before you begin
  4. Data requirements – Spell it out beforehand, ask (and answer) questions like: What’s the volume of data received? Who/what will be the recipient of this data? How do we retrieve it, process it, manage it, secure it?
  5. Structure is key – Use open, flexible and scalable tools that yield quick (but viable) results as opposed to monster-sized app or system
  6. Shop wisely – There’s no “one tool fits all” so carefully scan and evaluate the market and its many vendors, as they come in all shapes, sizes and prices
  7. Pilot please – Before you embark, it’s highly recommended to sponsor (pay for) and run a pilot project to build management’s confidence and validate your project strategy and roadmap

We’re of the opinion that big data analytics will have a great impact on the manufacturing world when it’s really put to good use, when it’s analyzed, not just collected and stored. You know, once, we didn’t understand data because we couldn’t see data. Nowadays, when we come head-on with vast amounts of data, we need to act fast, generating the ‘right’ data – the actionable data.

We need our data to drive growth, push performance and to no longer be seen as just another buzzword.


Share on facebook
Share on linkedin
Share on twitter
The Author
Efrat Nakibly
Director of Consulting, Priority Software U.S.

Sign-up for Priority news