ktsdesign - Fotolia
For today's manufacturing enterprise, data rules. This may be driven, in large part, by the prevalence of sensors and data pickup points in dispersed locations throughout the business enterprise. This pipeline of data has created a life of its own, often referred to as the Internet of Things (IoT).
The ability to assimilate such data via the Internet offers much opportunity to put this data to work by creating big data applications that incorporate the right technology and best practices to bring forth excellence in manufacturing.
Internet of Things gains momentum
"The Internet of Things is one to watch," said Paul Nashawaty, director of product marketing and strategy for Progress, a software development company based in North Carolina. "IoT is not a new concept, but it has been gaining momentum, especially for industrial firms. In the past, industrial firms have had many different ways to catalog and distinguish data pulled from different legacy applications without standardizing around one big data source."
"As it stands now," Nashawaty said, "firms are starting to migrate to new technology to help future-proof their environments and need a 'connector-tolerant' application to ensure connectivity with a multitude of sources such as MongoDB, SparkSQL, Hadoop, etc. Although the industry as a whole is headed in the right direction, it is still fragmented and lags behind other industries' efforts to utilize the new technology in big data. Those firms that do not embrace this evolution run the risk of being phased out."
Embracing this change, however, requires some safeguard to ensure that big data collection and assimilation becomes a help and not a hindrance. Steve Ehrlich, senior vice president of marketing and product management at Space-Time Insight in San Mateo, CA, points out that raw data needs to be channeled into meaningful information.
"The tendency of business users is to ask to see all the data that is being generated by sensors, assets and others," Ehrlich said. "Not only is this often impractical from a volume, performance and infrastructure perspective, but it is also not the best way to think about big data since it is impossible for humans to interpret and find meaning in so much data. The solution is that analysis of the data should go hand-in-hand with its collection and storage, allowing software to look for the anomalies and significant details in the data and presenting just those to users. This focuses businesses on the opportunities and issues they should care about and frees up personnel from manually collecting and correlating data themselves. A change in organizational mindset and processes is often required to make this a reality, making sure teams have access to the right software and data plus the training and skills needed to embed analytics into the way the organization does business."
Big data applications transform IoT into useful information
Assimilating big data and transforming it into usable strategic information is part of the burden that organizations take on to yield the fruit that big data produces. It entails more than a massive data fiesta arriving in one supercomputer from all sources. The quality and accuracy of such data needs to be well-managed to put it to use.
Manish Sood, CEO of Silicon Valley-based Reltio, emphasized the importance of data reliability and quality. "While NoSQL allows you to just dump information into a data lake, doing so will just cause issues down the road," he said. "Keeping master data management (MDM) as a separate discipline means that the silos of data within the organization, including packaged applications -- each with their own versions of people, product, organizations and other data profiles -- need to be accurate before being combined with big data from other sources. MDM rigor needs to be applied regardless of big data volume."
With the right rigor, good data provides value. If all efforts are focused properly, there's opportunity for manufacturers and industrial firms to hit notable home runs.
Firm hits home run with big data application
Manish Gupta is chief marketing officer of Liaison Technologies, headquartered in Atlanta; the firm consults on data utilization to enterprise companies, including the manufacturing sector. Gupta described one of his firm's home runs.
A company used big data to expedite the drug discovery process to bring new medicines to market faster. "Using our system, a prominent pharmaceutical company was able to source data from over a hundred clinical trial sites around the world in a variety of forms -- from handwritten documents in Nigeria and faxed materials in Japanese to electronic data from London," he explained. "We were able to handle all integration, persistence, harmonization and finally exposing APIs to a BI tool from our cloud-based platform. This approach led to insights faster than would have been possible with alternative methods, and with more accurate, reliable results. Net result: a shorter clinical trial process that helps move new drugs to market faster."
Big data project reduces online data residency
In another example, Dr. Werner Hopf, CEO of Dolphin Enterprise Solutions Corporation in Malvern, PA, explained that an international bottling operation for a leading beverage manufacturer was tasked to help reduce the total cost of ownership of its ERP system, SAP ECC.
"The organization has archived more than 11 terabytes of data, which has facilitated the consolidation of international operations, a major replatforming of their systems, and led to an overall increase in reporting performance." His firm implemented a phased approach across both its ECC and BI systems. It also required a strategy that could accommodate two very different business models.
As a result, the project reduced online data residency periods to just six months, implemented automatic archiving to allow the organization to catch up on historic archiving and meet targeted residency periods, and resulted in faster system performance, backups, upgrades and recovery times. "SAP ECC data was reduced from 32TB to just 10GB, which increased job performance. It helped modernize plant maintenance systems and improve the performance of the business intelligence report," Hopf said. "Lastly, it created an easy separation when the organization decided to spin off one of its business units into a separate entity."
Ravi Bapna, the Curtis L. Carlson Chair of Business Analytics and Information Systems at the University of Minnesota in Minneapolis, said a new post-mobile wave of digitization is on the rise and is comprising smarter physical ecosystems, sharing economy-based services and dynamic decision-making. "Tomorrow's smarter physical ecosystems are going to arise from the ongoing Internet of Things movement," Bapna said, "which essentially boils down to using sensors to connect physical objects such as machines, homes, automobiles, roads, even garbage bins and street lights, to digitally optimized systems of production planning and control and governance."
Learn about industrial IoT trends
Read an opinion about IoT hype
Get an update on IoT standards