Is ‘Big Data’ Too Big to Handle?

Is Big Data Too Big to Handle?

My favorite definition of Big Data comes from Gartner: it’s a combination of high volume, high velocity and high variety information.

Walmart is a great example of high volume, tracking one million transactions per hour. FICO is not only storing data from 2.1 billion accounts but monitoring it real-time to protect its customers against fraud.

And Facebook – is there any better example of high variety, with 2.1 billion users and 500 billion photos?

Do you think you need big data? Well, you can actually can get lot of value from little or just right data. You don’t need to spend millions of dollars to get it; rather just enough to gain value. Big data is not the always the answer but it is important to understand the concepts around it because those concepts can apply to any manufacturer. The big data advantage lies within analytics capability to draw conclusions to identify patterns.

So it’s really not about BIG data … and it’s not about ZERO data. It’s about data that’s JUST RIGHT.

With such a variety of structured and unstructured information you could capture, one of the best ways to get moving in the right direction is to start with the most critical asset or process causing problems, and then envision the end state – or what you want to achieve.

Two questions will help you get started:

  1. What data do I need to collect (pertaining to this asset or process)?
  2. What could that data tell me if I had it?

By focusing on what’s practical for your operations, you can take the overwhelming concept of Big Data and turn it into what I like to think of as the ‘Just Right Amount of Data.’

There’s good reason to scale your efforts. The power of Big Data isn’t in the amount of data you’re collecting; it’s how you use the data that matters.

There’s an acronym that will help you know what data you want to collect: STAR

  • Simple is get to the basics of allowing people to make quick informed decisions by providing clear visuals.
  • Timely is also basic, with today’s connected enterprise, its easier to get to the data quickly, so make sure you are looking at timely information.
  • Accuracy is important, especially for trust and improving your culture of decision making.
  • Relevant, goes along with the others, basic but important, if you are being measured on profit, you better have metrics to help you, if its more about uptime, you better have a KPI (key performance indicator) trained on that

I know, with new technology and new capabilities, the urge is to collect everything. The pendulum is swinging from making decisions based on a hunch to collecting data for the sake of having data. I advocate a spot in the middle.

Finding the ‘Just Right Amount of Data’ is a process:

  • Learn about the concepts. It is easy to get lost in Big Data. Focus your efforts on the most troublesome areas.
  • Plan your data collection. If your culture isn’t ready for the shift to data-dependence, then figure out where, and from whom, you need support.
  • Ensure data visibility. Once you have the data, you need to share it and empower people to make decisions that bring more success to your organization.

And remember: the value of Big Data is not in the numbers or statistics or reports; the value is what you learn from the analysis of the information. That means more data is not necessarily better. Some data is better than none, but getting the right data to right people to make the right decision is the very best scenario.

Chirayu Shah
投稿日 2014-10-13 投稿者 Chirayu Shah, Product Manager, Information Software, Rockwell Automation
  • お問い合わせ:

お問い合わせ

ロックウェル・オートメーションならびに当社パートナの仕事は優れた知識を通じて、自動化の設計と導入をお手伝いしその投資がいかされるようサポートすることです。

最新ニュース

ロックウェル・オートメーションにご登録いただければ、最新のニュースと情報を電子メールにて直接お届けします。