Research & Innovation
Manufacturing analytics: harnessing data to streamline operations
By Pulp & Paper Canada
By Pulp & Paper Canada
It’s a scary word to those less-than-tech-savvy business owners and operators out there, and particularly for manufacturers who do most of their analysis on the shop floor.
But harnessing data and using it to streamline operations is easier than you think, according to the gurus at Deloitte.
And it’s even easier for small to medium-sized enterprises that can take advantage of an abundance of free tools to maximize budgets and productivity for little to no cost at all.
According to Rhys Morgan, senior manager in technology strategy with Deloitte’s Edmonton office, it’s called ‘the scale paradox’; the idea that being a smaller-sized firm makes it easier to invest in the power of analytics.
“What we’re finding is that the smaller organizations with the least capital are sometimes better positioned to leverage analytics as a concept,” he said.
“The reason for that is they are smaller (and) this work is less capital intensive. The bigger organizations that have made more significant investments (elsewhere) may not be as well positioned as some of the smaller organizations.”
The first hurdle to get over when considering how analytics can help you trim the operational fat is getting over the word itself, Morgan said.
Speaking at the Western Manufacturing Technology Show (WMTS) in Edmonton, Morgan said the best way to approach the idea of analytics is by simplifying it.
Instead, look at it as simply extracting information from data—data you probably already have, Morgan said.
Generally speaking, there are three sets, or streams, of data to break down: internal; external; and extended value chain.
“There’s a significant amount of data being created across the (manufacturing) sector from multiple sources,” Morgan said.
Internal data—sourcing, production logistics and sales—is very diverse, he said, but is infrequent and not the greatest quality to work from.
Extended value chain data—supplier, distributor and customer—isn’t widely used yet and therefore can be difficult to track, while external data—economic, geospacial, statistics, mobility and social media—is often traceable through free channels.
“You need to look at every single bit of data that you have, look at what the sources are, look at how you’re gathering it, look at the quality of it,” Morgan said.
Whether it’s third-party supplier data or the information found within the piles of paper on your desk, the problems you can solve using analytics include trend analysis and causality; operational efficiency; cost optimization; and product management, to name a few.
Predictive analytics—using the past to predict the future—can even be done using existing data to do a simple-yet-effective workplace injury analysis, potentially saving lost time and money.
According to Morgan, analytics can even help you go lean by breaking down down issues and inefficiencies that have historically stood in your way.
“Having true insight over your data will help you identify those areas that aren’t lean (and) will help you make better decisions around those areas that may need to be leaned, processes that need to be optimized,” Morgan said. “Then you can target your lean agenda.
“Everyone’s making big investments in lean manufacturing, but where are you going to start? If you have a better understanding of your organization you know where to invest your lean dollar much more efficiently to get a better bang for your buck.”
And the best part is, applying analytics to your business is only as expensive was you want it to be, according to Morgan.
While hiring someone who can accurately analyze data may not be realistic for small capital manufacturers, having the right skills and tools is sometimes all you need to get started down the path to streamlining your business.
“It’s a very easy thing to do with the right focus and the right skill set and the right effort,” Morgan said. “What comes out of that is a significant amount of value in terms of getting true insight (out of) data.”