Operationalizing Big Data to Improve Quality and Reduce Costs
As Big Data and the Internet of Things (IoT) continue to reshape the competitive landscape of industrial manufacturing, converting information into intelligence has become critical, and manufacturers are seeking new ways to streamline production, optimize resources and reduce overhead expenditures.
“Big Data solutions can drive improvements and efficiencies in manufacturing over a variety of quality-related applications.”
The manufacturing industry generates more data than any other sector of the economy, and exponential data growth is guaranteed to continue. Manufacturers integrate data from sources daily, like device sensors, production systems, supply chain/inventory management platforms, sales forecasts and point of sale systems in order to extract insight to improve business processes and control costs.
As devices and systems become increasingly connected and intelligent, Big Data continues to compound, enabling manufacturers to increasingly invest in technologies to collect, organize and securely store their growing data volumes. According to a recent report from Technavio, the market for Big Data solutions in the manufacturing sector will reach $12.23 billion by 2020.
However, a key challenge for manufacturers is gathering, analyzing and acting on data quickly enough to deliver value to operations. For many, data analysis is still used to react to problems that arise rather than preventing problems before they occur. Transformation is required to not only manage the real-time data flow, but also better convert data into intelligence so issues across the production line can be proactively identified and addressed.
Manufacturing executives know all too well the operational and financial incentives for leveraging Big Data to improve quality assurance procedures. The American Society for Quality has estimated that many organizations have quality-related costs which account for 15 to 20 percent of sales revenue, with some going as high as 40 percent of total operations.
Big Data solutions can drive improvements and efficiencies in manufacturing over a variety of quality-related applications:
• Real-time analytics can be evaluated across work-in-progress products to pinpoint manufacturing defects early in the production process and forecast potential process or design flaws.
• Analysis of historical sensor data can reveal subtle anomalies indicating product flaws and provide early-warning analytics that correlate real-time measurements with quality models.
• Advanced analytics can reveal quality and performance variances down to the machine or operator level, allowing for adjustments to production processes on-the-fly to boost quality while cutting costs.
A Big Data analytics engine like Hadoop can be incorporated into a manufacturer’s infrastructure to aggregate data from multiple sources throughout the plant, and produce data-driven insights to help them improve product design and quality, as well as identify potential problems in the production process.
Energy costs represent another significant area of spends for manufacturers, and industrial companies continuously look for ways to reduce energy consumption and eliminate waste. The industrial sector accounts for approximately 31 percent of all energy consumption in the U.S.–that’s over 21,000 trillion Btu consumed each year–and much of this energy is used for manufacturing processes.
However, energy use can also be hard to track and quantify. Particularly for manufacturers, pinpointing which parts of the line are taking the lion’s share of energy can be a daunting task, one that requires visibility into the entire production process.
Manufacturers have access to an enormous volume and variety of data which can hold the solution to this issue. But many are not able to effectively measure energy metrics, in part because they don’t know how to handle the data flowing in from disparate systems and sources, let alone how to sift through and convert it into waste reduction, improved operational efficiencies and increased profits.
Using Big Data solutions, manufacturers can adopt a more strategic approach to energy management and move away from siloed data management techniques of the past. Advanced analytics and modeling techniques offer the potential to help decrease energy usage and drive operational efficiencies in a number of ways:
• Equipment sensors can track energy consumption patterns down to the device level for visibility into actual energy usage and revealing wasteful areas.
• Real-time energy profiles can help optimize production scheduling by cutting machine idling times back and increasing production on devices which are more efficient.
• By tracking anomalies, they can identify less efficient machines to target repair or replacement.
“Smart” plants, where all devices and systems become interconnected, are crucial to manufacturers who want, and need, to act upon 100 percent of their data. A recent IDC report confirmed that smart manufacturing programs can deliver financial benefits that are tangible and auditable, and tie directly into metrics such as revenue, costs, and asset levels to justify investment for manufacturing companies.
An Internet of Things approach, one that allows all systems to collect and exchange data regarding current status, operations and environment, can help manufacturers dramatically increase efficiency and enhance data-driven decision-making. Data should then be aggregated to a central location where trends can be defined and energy usage monitored in real-time.
Big Data software platforms can be integrated into existing IT environments, to extend analytic insights across all manufacturing data sources. Large IT vendors are available to help manufacturers transform their IT environments and maximize the value of business data.
Manufacturers have a huge opportunity to benefit from their Big Data insights. Those that are able to leverage data to achieve real-time, plant-wide intelligence will be able to make the data-driven decisions needed to succeed in a highly competitive climate.