Advanced Computing in the Age of AI|Thursday, August 13, 2020
  • Subscribe to EnterpriseAI Weekly Updates:  Subscribe by email

The New ‘King of Cool’ in Manufacturing 

Recognizing their competitive value, "Big Industry" has always pioneered the development and use of analytical tools. The key word, though, is "big." Unfortunately, the advantages of analysis tools such as finite element methods have been sorely missed by the small- and medium-sized manufacturers (SMMs) in the United States.

Recognizing their competitive value, "Big Industry" has always pioneered the development and use of analytical tools. As far back as 1965, when NASA issued a request for proposals for the development of finite element software, large companies pounced on the opportunity to advance the ability to gather and analyze data. The key word, though, is "big:" BIG industry has pioneered development and use of analytical tools.

NASA FEM logoUnfortunately, the advantages of analysis tools such as finite element methods have been sorely missed by the small- and medium-sized manufacturers (SMMs) in the United States. Limited access to resources, knowledge, and specialized software have hobbled the adoption of advanced analytics among industry without the "Big" prefix. Recognizing these challenges, NCMS has embarked on a program to "democratize" advanced modeling and simulation by spearheading the creation of Predictive Innovation Centers (PICs). PICs will grant access to high-performance computing (HPC) to businesses of all sizes by establishing a collaborative, time-share environment in which users face no initial investment and do not require highly trained personnel to interpret analytical data.

In addition to HPC, a new wave of analytical tools — tools of decidedly mixed parentage — have emerged in industry over the past decade. While many have originated from the classic stereotype of "boring" statistics, others have grown from what could arguably be called the new "king of cool" (or geekdom!) — computer science. The melding of traditional statistics and modern computer science has created a new set of tools with the ability to devour the data explosion and collect information from every user mouse click. Those who wield these tools have enormous power to predict which way markets or customers are trending. It is an entirely new field of data manipulation and comprehension, labeled with a variety of modern-sounding monikers: business analytics, machine learning, data mining, predictive analytics.

Strangely enough, Big Industry seems to have largely overlooked the trend, with the exception of some conglomerates such as GE which have gained solid expertise in analytics from their non-engineering domains. For the SMMs, the challenges to implementing analytics are unsurprisingly similar to those of HPC: investment, software, and expertise.

NCMS member company SimaFore recently concluded an important survey on the subject of business analytics, and discovered that fewer than 25 percent of SMM executives were aware of the potential business analytics offer to their operations. On the flip side, when stripped of jargon, more than 60 percent of manufacturing companies fully recognized the value of affordable software "apps," tools that address specific business problems such as accurate cost estimation, customer retention or predictive maintenance of equipment.

So what exactly is business analytics? How can small- and medium-sized businesses benefit from these tools? Analytics is about employing data to generate business insights to make objective and timely decisions. As long as the basic requirement for deploying analytics, viz. cheap and accessible data, is satisfied there are many ways it can add value to SMMs.

A typical challenge that is faced by most small businesses is growing their top line while at the same time controlling their marketing expense. One way to accomplish this is by gaining insight into their existing customers' buying patterns — identifying the demographic and buying profiles that fit, then using this information to predict where a prospect would likely fall. By using techniques such as clustering for example, businesses can quickly segment their customer database and then spend focused marketing efforts on only those prospects that fall into desirable clusters.

Another area where manufacturing companies can benefit is in operations, predicting time to failure of equipment by analyzing utilization and operating environment data. Neural network models, for example can quickly learn from available data and predict if key equipment is likely to fail, providing advance notice for maintenance and upgrades.

Small companies don't need to be big to succeed; they just need big tools. HPC and business analytics are big tools; and in order to remain competitive, awareness of and access to them is mandatory. While politicians and even business leaders endlessly debate how to restore our economy, the actual approach is simple: keep our corporate infrastructure competitive on a global level by ensuring it has the tools it needs. Competitiveness in turn breeds growth and employment, which drive a robust economy.

Add a Comment

Do NOT follow this link or you will be banned from the site!
Share This