Three steps to becoming a data-driven organization

Companies looking to take advantage of Big Data can become a data-driven organization by being committed to the cause and deploy solutions to capture and store data and use advanced tools to interpret the information they're receiving.
By Richard Phillips, Michael Stephens September 27, 2017

A factory performance dashboard provides a real-time and historical digital view of the manufacturing operations. Courtesy: PolytronIncreasing amounts of data are being generated by manufacturing operations—from smarter devices and machines, tighter supply chain integration, and more—but for most manufacturers, hardly any of that information is being used to drive business decisions. As manufacturers begin to make better use of Big Data, they are realizing it represents both a big opportunity and a big shift from guesswork to data-driven.

Big Data will allow an organization to better track and react to factory performance in real time, including improve quality, reduce costs, increase responsiveness, improve planning and support Lean initiatives. Ultimately, big data will provide tighter integration, greater visibility and improved analytics that will drive manufacturing excellence and responsiveness. Early adopters will develop a competitive advantage as their workforce shifts to data-centric thinking. When becoming a data-driven organization, follow these three steps.

1. Commit to becoming a data-driven organization. Placing data at the heart of an organization is the foundational step in the process to becoming data-centric. According to IBM, there are four types of data: Volume, variety, velocity, and veracity. Volume is ensuring you have a sufficient amount of data (not a problem for manufacturers). Variety means the different types of data being collected such as counts, measurements, binary, or nominal. Velocity is the amount of data needed to paint a complete picture. Finally, veracity is the certainty around the collection of your data. Automating data collection can increase all four V’s. Once collected, the data needs to be converted to information within context and provided to the right person (or system) at the right time.

2. Deploy solutions to capture, store, and structure data. With your new commitment to being data-driven, it’s time to talk about capturing, processing, securing, integrating and contextualizing the data. Thousands of sensors sending data each millisecond through a smart manufacturing infrastructure can overwhelm even the most robust network. Careful planning and clever data collection methods can help reduce network traffic. The user requirements should dictate the data required so that any data collected will have a use. Once you have decided on how you are going to capture and structure your data it’s time to analyze this data.

3. Add data analysis skills and tools. This is where the rubber meets the road. To make the most of becoming data-centric you need to ensure the foundation is set. This means collecting the right data. Once you have verified the data required, you can start the path of analyzing it and using it to impact the future of your organization. To do this means focusing on technologies that can give you the upper hand.

Pick the right manufacturing enterprise software (MES)

In today’s competitive market it is about doing more with less. Process improvement initiatives aim to gain a competitive advantage by collecting, analyzing, and correlating data to determine root causes of process bottlenecks and inefficiencies. The right MES is required in today’s complex business environment of continuous change because it enables business agility by accommodating unforeseen changes and automating the information flow.

When selecting an MES, be sure to look for some key features. First, the MES should run on proven technology. It should also be an integrated platform where new features can be activated without requiring new software. Lastly, it should come with preconfigured functionality to help you get up and running quickly for many of the more popular applications, such as OEE, SPC, and e-records.

There are various disciplines in manufacturing that need to come together at the same time, and a higher automation level is required for the separate systems to function efficiently. The right type of MES can help you automate the data-collection and information flow, turning process improvement concepts and practices into decision support tools directly tied to the overall goals and objectives of the business.

Richard Phillips, PE, PMP, product lead for manufacturing intelligence group, Polytron, Inc.; Michael Stephens, executive vice president, customer service, Parsec Automation Corp. Edited by Chris Vavra, production editor, Control Engineering, CFE Media, cvavra@cfemedia.com.

ONLINE extra

About the authors

Richard W. Phillips, PE, PMP, Polytron, Inc. manufacturing intelligence group. Phillips’s expertise and focus on strategic plant floor data management provides solutions to enable manufacturers to execute data-driven decisions across the enterprise—from supply chain to production to the warehouse. He has consulted and implemented smart manufacturing strategies for major manufacturers to help provide a roadmap for effective plant wide data management. Phillips has over 24 years implementing capital projects with a focus on packaging, process, and material handling systems. He holds an M.S. and B.S. in electrical engineering from Auburn University. Phillips is a published author and speaker for manufacturing intelligence and smart manufacturing related content.

Michael Stephens, EVP, customer Service, Parsec Automation Corp. With three decades of experience in Big Data, analytics, visualization and, more recently, IIoT, Stephens leads the Parsec customer service team in delivering significant business value through the design and deployment of software solutions for data-driven manufacturing.

Polytron, Inc. and Parsec Automation Corp. are CSIA members as of 9/27/17.