Noodle.ai CEO: Mills That Adopt AI Will See Returns. But the Size and Speed of the Payoff Depends.
05/19/2020 - There are dividends to be had in applying artificial intelligence (AI) to steel mill operations, but the pace and degree of those returns depends on the operation to which the technology is applied, according to Stephen Pratt, chief executive of data science and artificial intelligence firm Noodle.ai.
"The fastest (path to a return on investment) is probably not the highest,” Pratt said Tuesday, delivering the first presentation in AIST’s new webinar series, “Digitalization and the Future of the Steel Industry.”
The free series, which will run throughout this summer, brings together a number of experts who will discuss of a variety of aspects related to the steel industry’s digitalization efforts.
Pratt told the audience that one path to a quick return on investment lies in asset health and AI that can detect operating anomalies signaling potential problems. That, he said, can greatly reduce unplanned downtime.
From there, a relatively easy second step is to apply AI to product quality. Product quality correlates with asset health data, so once the hard work of installing data pipelines for anomaly detection is done, there’s not a big technical leap to product quality.
On the other end of the range, the big returns for steel mills are to be had in AI that analyzes materials mix and is capable of optimizing those costs, leading to dramatic improvements. And at the pinnacle, he said, is AI that schedules production based on profit per mill hour. However, that’s very challenging, he said, because of the variability and permutations in production from mill to mill.
“That is something that takes quite a long time to do because every mill is very different,” he said.
The next installment in the series takes place 10 a.m. Wednesday with Katja Windt, chief digital officer for SMS group. SMS group and Noodle together launched an AI-based application specifically for the steel industry.
You can register for the webinar here.
If you can’t attend, check back here later this week for a replay of the presentations.
The free series, which will run throughout this summer, brings together a number of experts who will discuss of a variety of aspects related to the steel industry’s digitalization efforts.
Pratt told the audience that one path to a quick return on investment lies in asset health and AI that can detect operating anomalies signaling potential problems. That, he said, can greatly reduce unplanned downtime.
From there, a relatively easy second step is to apply AI to product quality. Product quality correlates with asset health data, so once the hard work of installing data pipelines for anomaly detection is done, there’s not a big technical leap to product quality.
On the other end of the range, the big returns for steel mills are to be had in AI that analyzes materials mix and is capable of optimizing those costs, leading to dramatic improvements. And at the pinnacle, he said, is AI that schedules production based on profit per mill hour. However, that’s very challenging, he said, because of the variability and permutations in production from mill to mill.
“That is something that takes quite a long time to do because every mill is very different,” he said.
The next installment in the series takes place 10 a.m. Wednesday with Katja Windt, chief digital officer for SMS group. SMS group and Noodle together launched an AI-based application specifically for the steel industry.
You can register for the webinar here.
If you can’t attend, check back here later this week for a replay of the presentations.
