100,000+
metal products manufactured
125,000+
customers
Fortune 500
company
EXECUTIVE SUMMARY
Reliance Steel teamed up with TensorIoT to connect its industrial machines to the internet and get data from sensors and loggers on the machines.
GOAL
How do you bring legacy industrial equipment online, to learn how to keep them running smoothly and efficiently?
RESULTS
We hooked up Reliance’s machines to the internet by adding sensors, and then ran machine learning models to make actionable predictive maintenance suggestions.
Background
Reliance Steel is the largest metal service center operator in North America. Many of the machines that Reliance Steel uses to produce goods and materials have been in service for decades. When one of them fails or breaks down, it lowers production quotas, loses revenue, and increases maintenance costs. By adopting a more preventative approach to maintenance using Industrial IoT and machine learning, Reliance could drastically reduce the down-time of broken machinery and possibly eliminate the production gap that offline machines can create.
The Challenge
Getting machinery online for one of the largest metal manufacturers in North America was no small feat. In order to prevent any disruption in operations and output, the team would need to work overnight to install sensor hardware. Then, with the sensors in place, the team would need to analyze the data. Given the large number of machines being monitored, there would be a ton of data that would need to be sifted through and analyzed. Once analyzed, the team would also need to craft a custom UI that could turn the large quantity of data into easily digestible visualizations that provided users with actionable insight.
The Solution
The first step in this project was to get Reliance’s industrial machines hooked up to sensors that could monitor and measure different states. After the sensors were in place, we ingested the data using IoT Core, and stored the data in a data lake, so it was available to use for machine learning and for a custom web UI. We then used Amazon SageMaker to run machine learning models that could detect when machines were about to break, so that preventative maintenance could be performed, and the machines could keep running.