Data is hard to use – at least in its current form across the industrial landscape. Operations data is siloed in different systems at the plant in and in corporate networks and in the cloud. It’s also fragmented in different formats, with varying quality and often lacking the context necessary for analytics. Studies show that anywhere from 80%-95% of companies’ data goes unused. This has a significant negative impact on strategic decision-making, decreases productivity, and results in hundreds of millions of dollars in lost value.
The typical data scientist may spend just 20% to 25% of their time actually using data to identify operational improvements. A recent report from Anaconda, a data science platform company, The State of Data Science 2020, finds that nearly 50% of a data scientist’s time is spent on data preparation—data loading and cleansing—and they lose even more time when contending with numerous data environments and dependencies.
The report continues to state: “As a result of these production struggles, fewer than half (48%) of <data scientists> feel they can demonstrate the impact of data science on business outcomes.”
Imagine if we could free up the time that our people spend wrangling data and allow them to actually affect operational productivity?
To quote a Digital Manufacturing Expert at a leading Building Materials Manufacturer:
“There is more than $1M in any manufacturing plant in opportunities to optimize productivity. Not having the right data holds you back from your ability to run at the speed you want to run at.”
By removing roadblocks around data access, management and environments, suddenly you've got four times the engineering resources you have today without having to add a single new person. Your team can now focus the majority of their time and effort on value-added activities.
People who are empowered with meaningful operations data drive business outcomes.
Consider this example of how one manufacturing plant, among hundreds of plants producing building materials, started to identify the right data and data model to inform their plant-level optimization efforts.
Using Element Unify as its platform, the company was able to take disparate metadata to create a standardized tags and asset hierarchy. Standardization of naming tags in turn enabled them to more accurately see what their assets and equipment looked like, build reliable data models and algorithms, and develop more prescriptive analytics.
With the right data models in place, their engineers now had the time to uncover opportunities to increase productivity and deliver on optimization efforts, while also decreasing costs.
During Phase One of the project, Element helped the organization achieve more than $1M in productivity gains at the first group of plants that participated in the initiative. Productivity gains came from a variety of sources including:
Productivity gains from just one plant offered the company a 50:1 return.
Until we move to a new model for managing and using operations data, engineers and data scientists will continue to waste too much of their time digging out data and stitching it together in order to build their case for operational improvements.
Questions? Please contact us.