Big Data

It's time to devise a plan to store all the RFID data you're collecting and a strategy to use it—in combination with other information—to gain insights that could transform your business and provide a competitive advantage.
Published: August 25, 2015

Big data. It’s a popular term characterized by the three Vs: the volume, variety and velocity at which companies collect data that flows from traditional databases, financial transactions, weather patterns, point-of-sale (POS) terminals, beacons, sensors, clickstreams, social media, log files and myriad other sources. Now it’s time to add radio frequency identification data to the mix, as organizations track and manage assets, inventory, tools and other things and people in real or near-real time—across supply chains and within factories, stores, hospitals, construction sites, stadiums and other venues.

Already, companies that have deployed RFID solutions are collecting accurate data automatically. They are dramatically reducing—if not eliminating—errors that result from manual tracking, including the use of bar codes. These organizations have gained visibility into the what, when, where and why of business processes, so they can turn a basic tracking tool into a business-intelligence tool and even a predictive analytics solution that can help them understand events and boost operational efficiencies and productivity while reducing costs and improving customer service.

Illustration: iStockphoto

Yet, beyond the return on investment that accrues from individual RFID deployments lies far greater business value. The ability to plug in and connect RFID data to other data sources for analysis can determine whether a company winds up on the leading edge and achieves a competitive advantage or lags behind. “Big data creates opportunities that can redefine and reinvent an organization,” says Scott Schlesinger, data and information management leader for the Americas at EY IT Advisory.

Consider, for example, a food manufacturer or pharmaceutical firm that is RFID-tagging perishable goods and tracking them through the supply chain. If the product is monitored with a temperature sensor, the company can also receive alerts if the item is exposed to conditions that fall outside acceptable parameters. If that RFID data is also combined with historical sales data, weather data, POS data and social- media data, both the manufacturer and logistics provider can get a broader and deeper understanding of demand and consumption patterns and use this information to optimize production and shipments, Schlesinger says.

Similarly, big data can revolutionize interactions with customers. A department store retailer, for example, could use RFID to track shipments and products at an item level, POS data to monitor demand and social media sentiment data to understand consumer preferences. By combining this data, it’s possible to gain a holistic view of the marketplace. It’s also possible to use beacons for real-time, highly customized promotions, says Bill Hardgrave, dean of the Harbert College of Business at Auburn University and founder of the RFID Lab. “A business is able to adopt an omnichannel and highly personalized approach that takes customer interactions to an entirely different level,” he says.

Industry experts agree that in the near future RFID and other big data will play an essential role in business, by making new insights possible across a wide swath of industries. “Big data is now at the center of everything,” says Antonella Mei Pochtler, a senior partner and managing director at Boston Consulting Group (BCG).

Big data “slides the dial” from a reactive mode based on reports and data after the fact to a proactive mode that uses software algorithms and predictive analytics to make better decisions, says Mark Beyer, research VP at consulting firm Gartner. In this environment, there’s a need for new thinking and new skills. Organizations must identify critical data and how elements intersect, tie together data sets, and break down silos that prevent them from achieving maximum returns.

Gaining Greater Visibility
RFID, which takes the errors out of manual processes, generates high-quality data that provides better visibility into processes and enables companies to make smart business decisions. Airbus Group, for example, increasingly uses RFID to capture data that managers then use to maximize efficiencies at the jet maker’s manufacturing facilities.

“RFID tells us what is going on in the physical world beyond what humans can track and measure on their own—it bridges the gap between the physical and IT worlds,” says Carlo K. Nizam, head of digitalization within the Information Communication and Technology team at Airbus Group. “The ability to connect objects and systems provides visibility into complex processes. This extra ‘real-world’ connectivity leads to an explosion of data that can have enormous value, especially when combined with big-data analytics.”

Recently, the airplane manufacturer began combining all that RFID data with other data sources to enhance visibility and gain new insights. When the company connected RFID data from different types of tools used in manufacturing plants with data from enterprise resource planning (ERP) and maintenance systems, it began to assemble a much clearer picture of which tools were used where, when and how within the facilities. Then, using analytics capabilities, Airbus identified optimal locations for tool stores that would reduce waste and more accurately predict when specific tools need calibration or replacement based on actual use cycles (rather than time-based predictions), and which tools were being used more and less frequently. This type of insight was unattainable with paper-based and/or manual “connectivity,” Nizam says.

Mercy health system, which operates 46 acute-care and specialty heart, children’s, orthopedic and rehabilitation hospitals, as well as nearly 700 other clinics and facilities, in Arkansas, Kansas, Missouri and Oklahoma, is focused on building a next-generation architecture that supports big data. Mercy is using an active RFID real-time location system to monitor equipment, supplies, staff and patients. Mercy is now exploring how to overlay RTLS data with other data sources—patient treatment information, electronic health record data, mortality rate statistics and more—to improve service delivery, medical treatment capabilities and overall performance, says Scott Richert, VP of infrastructure. “It’s one thing to display a map of the floor and know where all your equipment is,” he says. “It’s another thing to understand how to buy equipment, when to maintain and replace it, and where it needs to be located to produce the best treatment and results.”

Gartner’s Beyer says a logistics and distribution company he worked with RFID-tags cargo but uses the technology for more than tracking the locations of pallets and other items. “They are using it to cube their trucks—they can determine how to best arrange items for the maximum load, as well as delivery efficiency prior to putting all the cargo on the truck,” Beyer explains. The firm achieves this by plugging in routing data and waypoint information, contract information that addresses penalties for delays, and traffic and weather data that could impact delivery times. Because the distributor knows exactly how the truck is cubed, it is possible to reorder and reload at waypoints, he explains. “They are able to generate a modeled representation using RFID,” he says. “The system takes into account a number of key variables, including fuel consumption, driver behavior, contract performance issues, penalties and taxation.”

Construction companies can also benefit from big RFID data, Beyer says. The operation of cranes, for example, might require maintenance every 25 days. But if you RFID-tag each crane for identification and input that data with weather conditions—humidity, for example—you can increase the safe operation of the crane. “We can say we have a potential situation here,” he says, “and we’re going to make sure we maintain it a little bit sooner than we normally do.”

Developing a Big RFID Data Strategy
At the heart of a big-data initiative is the ability to collect and assemble the right data and make sense of it. And like any RFID project, any big-data project must begin on the business side, says Ken Traub, president of Ken Traub Consulting. A company first needs to identify the business case. Then, IT can enable the new processes and capabilities.

Companies should begin to think about how they will develop a big-data strategy, Traub says, adding that you can’t just go out and hire a big RFID data specialist. Someone in-house—an individual or a team—needs to understand the RFID data you’re generating and how it relates to your business, he says. These people need to think creatively about the benefits of knowing where things were (historically) and are (in real time). Before you can begin to use analytics software, you need to know what questions you want answered.

Airbus’ big-data initiative focused on two primary goals: to make better use of existing data by tapping into separate systems, and to expose more granular operational data, Nizam says. “We asked, ‘What do we want to happen? What do we want to find out? What useful information can we generate and how can we put that to use?'” he explains. “A team then mapped out different attributes and information sources, including where the required data resided. The final step was connecting all the data sources to make the desired information accessible, actionable and meaningful at the touch of a button. At the center of all this was the enabler for real-world data: RFID.”

Although Nizam won’t disclose details about specific big-data projects across the company, he isn’t shy about affirming Airbus’ commitment to the technology. “Big data is a highly strategic topic—not only for the digital transformation initiative inside Airbus Group but also for each of the divisions in the group itself,” he states.

In retail, most stores deploy RFID to improve inventory management. The big-data component begins when retail executives ask, What do we want to know about our customers to improve sales? Companies can then develop systems that allow them to detect changes in consumer behavior and sentiment and make changes to stock more quickly, says Adebayo Onigbanjo, senior product manager at Zebra Technologies. It also might mean knowing a customer’s preferences and aligning e-mails and promotions to move stock more efficiently while better matching consumer preferences, he adds. The inventory piece remains important. “You still need to know your inventory counts,” he points out, “but you are able to use data and information in a multidimensional way and across channels and systems.”

It’s essential to hire and train employees who can bring the right set of analytics skills and data science to the enterprise, Onigbanjo says. Transforming data elements into meaningful results is no simple task. In many cases, it’s necessary to view the business and various processes from an “outside-in” perspective, and think beyond the narrower domain of a department or division, he says. Data experts can create transparency and determine what type of data to capture. They also recognize which devices, sensors and systems are required and how to connect them to the right algorithms and business logic. “As the situation becomes exponentially more complex, you need the right combination of tools and technologies,” he notes.

It’s also important for business leaders to look beyond the four walls of their organizations when assembling a big-data initiative. This often means establishing new partnerships and business arrangements that revolve around shared or pooled data. This can lead to more sophisticated marketing campaigns, improved training and hiring decisions, enhanced research and development, and better operational decisions, real estate purchases and other investments.

Putting the Pieces Together
RFID generates a lot of data, and many companies don’t keep it all—once they’ve tracked, say, shipments or inventory on a sales floor, they may assume it’s no longer useful.

Don’t dump your data, Traub advises. Even if your company is not ready to plan a big-data strategy, develop a software strategy for storing all your data. “There are many products on the market for managing large quantities of data,” he says. “They have different approaches, so you have to research them and decide which works best for you.”

Companies also should begin to investigate new tools that do calculations and outputs on an ongoing basis, to process data in real time, Traub says. These new “large-scale real-time event-processing tools” are still immature, he notes, but they will be a major component of a big-data strategy.

In addition, companies will need to build an IT framework that facilitates the exchange of data, says Su Doyle, senior marketing manager at Checkpoint Systems. This typically involves clouds and APIs that connect databases and systems.

EY’s Schlesinger says the goal isn’t to make different data sets available to different groups but, rather, to make the right data available from the same central data repository. “Data, including RFID data, can be used in entirely different ways to achieve completely different insights that help run the business better,” he explains.

To facilitate this type of environment, many organizations are turning to next-generation computing platforms such as Apache Hadoop, which offers a robust framework for storing data and running applications on computing clusters that rely on commodity hardware. The highly scalable open-source environment delivers massive processing power and vast storage. SAP HANA, which relies on an in-memory, column-oriented relational database management to process large volumes of data faster and more efficiently than traditional databases, is also gaining favor. Tying together these two technology solutions creates what Schlesinger calls an integrated platform that paves the way for enterprise analytics. What’s more, these systems also make it easier to use clouds and APIs to streamline data flow. “Without the right framework in place, the amount of data becomes untenable and it becomes impossible to get any value from all the data,” he explains.

Regardless of the specific technology, it’s critical to break down data silos and link data from legacy environments, including ERP and supply-chain management systems. Hadoop, for example, allows a business to process vast amounts of sensor data and contextualize it before it is slotted into a database. Yet, organizations must also look beyond technology to break down data silos. Too often, Zebra’s Onigbanjo says, RFID projects exist as islands. Because there’s no oversight at the senior executive level and no task force or team overseeing initiatives, groups within the enterprise don’t share existing knowledge, and organizations wind up duplicating efforts and losing data. “There must be executive support and a structure in place,” he points out.

Connecting to Results
For many businesses, a sophisticated big-data initiative may still be a few years away. But experts say now is the time to begin developing a strategy, building a big-data and analytics framework—and piloting projects. Digital technologies are advancing rapidly, and without a basic foundation and structure the task will be even more difficult in the months and years ahead.

As the business world transitions to an era where data is abundant, it’s vital to adopt a clear strategy, BCG’s Pochtler says. “A big-data initiative cannot be catch-as-catch-can in terms of simply collecting data and then trying to figure out what you can do with it. You have to focus on the core questions you want to answer and what type of data and analytics are necessary to produce the results that really matter.”

When organizations build a big- data framework, it’s possible to create a more agile and flexible business that’s better prepared for the challenges of the digital age, Hardgrave at Auburn University says. “By marrying a variety of data sources, a company can move from running the business based on past events to running the business on current and, with predictive analytics, future events.”

“The business world is undergoing a paradigm shift,” Airbus’ Nizam says. “There is a growing awareness that data is important, and it unlocks answers to complex questions. The next stage in the paradigm shift will be to ask, Where does data come from and how is it collected? This is where technologies like RFID will play an important role in building greater connectivity and insight into the industrial ‘Internet of Things.’ Together with big data-analytics, this has the potential to truly transform business and industry.”