With on-demand expertise and visual guidance delivered through AR technology, Harpak-ULMA can support customers from anywhere and introduce new capabilities like virtual factory acceptance tests. Meanwhile, customers can use the technology to reduce operator learning curves and on-boarding times.
Creating new insights
Harpak-ULMA has made significant progress on phase two, and is now underway on phase three, which involves incorporating more IoT touchpoints. This essentially combines packaging machines with their digital twins to offer enhanced operating experiences and information integration.
“Expanding the breadth and depth of embedded IoT enables us to integrate with advanced IoT applications that expose detailed machine operations in real time to production staff,” Roach said.
For customers, more data means more real-time, contextualized production information is available to both staff and onboard diagnostics.
“In this stage, we’re expanding on AR capabilities to deliver a deeper set of real-time capabilities that will allow staff to have what we call X-ray vision,” Roach said. “Exposing the digital twin of a machine is an interactive way to learn more and diagnose better.”
For example, if a technician gets an alert about a machine problem, they don’t need to shut down the machine, go through lockout/tagout procedures and open the control cabinet to investigate the problem. Instead, they can just use an AR headset, smartphone or tablet to look at a digital twin of the machine and identify the fault.
The fourth phase of Harpak-ULMA’s digital transformation is to apply machine learning and artificial intelligence to the vast amount of data collected during production. This effort, while still in the planning stage, will allow the company to create new business models by introducing cloud-based predictive maintenance and benchmarking services.
“This final stage represents an evolutionary step for customers in terms of their OEM relationship,” Roach said. “Big data solutions that leverage machine learning and artificial intelligence will make the holy grail of predictive maintenance analytics in our markets attainable.”
For customers, predictive maintenance can help them reduce unplanned downtime, as well as restructure legacy maintenance and cost models. And industry wide benchmarking offers them an opportunity to get a new understanding of the effectiveness of their production processes.
Of course, making these concepts a reality in customers’ production environments will come with challenges.
“We’ll need to resolve data ownership concerns and work through connectivity issues that arise with predictive maintenance monitoring,” Roach said. “But it’s not like there’s no precedent. If this can be done for aircraft engines, it can be done for a packaging production environment as well.