Recently I was talking about future opportunities in data analytics, which brought up these 2 examples. They are from the AME presentation by Andrew Au. If you have a chance to listen to him, do so, he’s dialed in to what’s going on. Here are his examples (thank you Andrew):
TetraPak
Swedish, support manufacturing in 175 countries, has invested heavily in cloud technology & holographic computing. One of the main challenges they look to solve for is “How do we avoid machine downtime?” obviously very costly downtime. They looked at specific equipment, servo motors for example used in forming & sealing for food packaging and said “How can we collect more data on these packaging lines so we know when we need to replace these components.” And particularly for the servo motor it has a lifespan of 2,000 – 7,000 hours. As you know the average factory is running for 4,000 hours a year. You’ve got up to a year delta in terms of when you need to replace these components. The challenge with the motor is if you don’t replace that motor in time, it’s going to destroy a lot of the other components. So they sensorized their packaging line; they took a 6 month pilot; 11 packaging lines; connected them to the cloud.
Within 6 months they could predict the failures within 5 of the 11 packaging lines. We take that a step further, they now have over 5,000 machines that are sensorized, collecting insight, and when their service engineers go to a customer site, they are now armed with that data, they’ve got it on their phones, insights from 5,000 different machines.
They are walking into the customer site much more equipped, much more knowledgeable; and to improve their first time fix rates they have equipped; they have a few hundred HoloLens’ that are now in the field, so when a service tech goes if it’s a machine they’re not that familiar with they can connect back to the service center and they’re being guided by a deep expert who knows a lot more about the machine than they do; and they’re remotely guiding them on what components they need to be testing.
An empowered front line and empowered field service operations.
JABIL. https://www.jabil.com/about-us.html
A few of the main headwinds they were seeing:
Our customers want mass customization
We’re seeing decreasing product life cycles
We’re seeing customers demand more & more data from us; they want insights
They’re on this journey evolving from being what they were known as is a contract manufacturer, to a full solution provider for manufacturing. One of their first steps similar to TetraPak is how do we sensorize our equipment & production lines, and by doing so they get all this data from their manufacturing processes, they look at supply chain they infuse that data, they get data from the field, and they’re applying machine learning algorithms to predict downtime for machines and also design flaws.
So they’re going back to their customers with insights of “you probably need to be tweaking this design this is where we predict we’re going to be running into failures 22% of the time.”
They’re shifting from a reactive to a predictive company as they evolve to a full service company. This is an example of how we can evolve to the customer need, which is ‘give me insights, give me stuff I can actually use, don’t tell me something I already know.’