Field Intelligence guest column
Small and medium-size enterprises often suffer from not having the personnel or resources to adopt Industry 4.0 solutions in their factories, meaning innovation falls by the wayside. In part because of this, smaller players can also sometimes have limited imagination when it comes to utilizing advanced technologies. So industrial IoT platforms haven’t traditionally targeted them.
Until recently, the cost of implementing a new data solution outweighed the clear return on investment, so solutions were mainly adopted by Fortune 100 and larger companies.
As of 2020, 90 percent of small manufacturers in the U.S. didn’t even have an IoT implementation plan. Without such a strategy, sufficiently trained personnel or infrastructure for data collection, manufacturers can’t make use of technologies. Consequently, it’s challenging to see the full picture of potential margin improvements even on tight budgets.
It’s been a year and a half since we started working with Daiwa Steel Tube Industries, in Tochigi, Japan. It is already the largest galvanized steel tube producer in Japan, but its president, Shin Nakamura, saw opportunities to improve efficiency and an immediate need to make up for capability lost due to impending worker retirements.
When it comes to implementing software, however, Daiwa doesn’t have a large IT department. It also does not employ data scientists; its most mathematically technical person is its manufacturing engineer. Shin engaged us to provide smart-factory tools and the specific analytics that bridged that resource gap.
Daiwa is emblematic of many smaller manufacturers around the world. Manufacturers will need to replace institutional knowledge lost by retiring Baby Boomers: Japan alone is set to face a shortage of 6.44 million workers in the labor force by 2030.
Utilizing advanced analytics—and getting new workers used to it—will be crucial for smaller players moving forward.
It will help to increase knowledge throughout the workforce, improve communication when issues occur on the production floor, and enhance visibility of key metrics.
From the outset, it was clear that Daiwa wasn’t collecting enough diversity of data for us to provide actionable insights.
For example, we could only access small samples of sensor data, such as whether the motors were out of range, but we couldn’t identify what impact that would have on the steel that they were producing.
We solved this by building our architecture to integrate sensor, enterprise IT, and streaming video all in one place.
Moreover, to gather data using other IIoT platforms, Daiwa needed to upgrade to newer generations of hardware that have open communication standards (such as OPC Unified Architecture).
Doing so brings additional operational risks, downtime and the cost of installing new hardware.
As a solution, we built our own drivers to bridge the gap to the older devices, ingested into a modern backend.
Daiwa operators can now interact with IndustrialML’s real-time dashboards, real-time alerts and have the ability to generate reports over any time period.
We have incorporated five different types of data sources to generate alerts, and we are actively tracking metrics on a week-to-week basis.
This has enabled the company to set improvement targets of 10 percent and higher for uptime and yield, potentially saving more than $1 million annually.