Rubber & Plastics News
April 12, 2019
By Chris Sweeney
As manufacturers move toward a more automated future, there are plenty of questions to work through, and a plethora of opportunities for those that can.
Automation, machine learning and what to do with all the data produced were some of the issues addressed by a panel of industry experts at the Plastics and Rubber in Automotive conference, held recently in Novi. The end goal for many companies is to create smart equipment, machines that can self-diagnose, self-correct and communicate.
“The idea is not just being reactive and finding out why something occurred on the prior shift,” said James Ricci, chief technology officer and co-owner of Harbour Results Inc., who also moderated the discussion. “We actually want to become more proactive and understand and anticipate issues, ultimately getting to something that’s totally actionable.”
Matt Myrand, director of advanced manufacturing and supply chain at Faurecia North America, an automotive supplier, said his company has a wide diversity of programs at very low volume, meaning it must adapt to customer volumes and products. What that could look like in an Industry 4.0 shop is machines flexing in and out based on the products being made.
Quick, reliable decisions are also a key driver for companies. As turnaround times in the automotive industry become more rapid, companies are turning to automation to keep up.
“One of the biggest drivers is cost and efficiencies,” said Shaun Karn, president of Hi-Tech Mold & Engineering Inc. and Baxter Enterprises L.L.C. “One of the other pieces to that is work force development. By utilizing this, we’ve seen a number of different ways how the flexibility can help us and we’ve seen how computers and machines can help drive decisions for you.”
Michelle Bockman, vice president of commercial development at HP Inc., added that the business drivers behind the movement involve customization and personalization.
“I would say in the last two years, we’re seeing the industrial market move to true industrial applications (for 3-D printing),” Bockman said. “They’re able to use parts in their sub-assemblies and in their vehicles. A lot of this is driven by cost and time to market. We’re able to make things a lot quicker that you couldn’t do before in traditional design or manufacturing.”
Karn said collecting data and using it helps predict pitfalls and make better decisions.
“There are a lot of different data aggregators entering the field,” said Mike McGrath, director of automotive and manufacturing at the SAS Institute. “The big challenge companies have is to take as much data as possible off of the machines, combine that information and ultimately glean results and predictive analytics.”
One key issue, especially for manufacturers who have been around for a while, is how to integrate these new features when the shop is still running older equipment.
Myrand said his company has a mix of machines that are decades old and others that have been purchased within the last year. There are both environmental and vibration sensors that can be externally mounted to the older machines, allowing them to gather a sizable amount of data.
McGrath echoed those sentiments, saying that while the data might not be as much as a newer machine, it’s enough for manufacturers to get a start and capitalize on the advantages of the new technology.
“Start somewhere,” McGrath said. “I think everyone in this room would have certain issues where you could potentially improve yield, ultimately reducing scrap and those kind of things. You know what your pet projects are, I would just start with something. Find a project, start small, start collecting data. There are a lot of cloud-based systems you can leverage to grab analytics tools on demand that are very user-friendly.”
Karn said his company has a new facility in South Carolina, and most of those machines have the technology and systems needed for automation and data harvesting embedded. But the company has to be more selective when it comes to the older machines used in Tennessee and Michigan, focusing on items that are critical issues. He often finds once a company starts automating the data gathered leads to other things that could be harnessed and things usually snowball from there.
“I think the only issue with older machines is you’ll be able to collect the data, but getting those machines to automatically self-correct may be a bit of a challenge based on technology,” Karn said.
For rubber product manufacturers, there are additional challenges beyond the equipment.
First, in certain high-quality driven industries like automotive, medical and aerospace, any time there is a change to a manufacturing process there are hurdles around customer approvals.
Jim Fitzgerald-CEO of Flexan LLC, a contract manufacturer of rubber, thermoplastic and silicone parts based in the Chicago area-said in a separate interview that automation for rubber companies is more challenging than in plastics because they’re dealing with a flexible part. This means part removal or extraction requires different tactics.
“The first challenge is not automating for automation’s sake, but automating because it’s a better way to do the work that you want to do,” Fitzgerald said. “I think you can fall in love with this idea of having an Industry 4.0 facility, but when you look back at it you have to ask yourself was the cost of the investment worth the benefit. You have to be practical about what you’re trying to automate.”
Rey Obnamia, vice president of technology and regulatory at IRP Medical, the silicone manufacturing subsidiary of the IRP Group focused on the medical industry, said in an interview that his firm already is starting to automate with the addition of robots. The focus is trying to make redundant tasks more efficient with machine learning to give the company a more competitive advantage.
“We’re trying to find a means of applying automation,” Obnamia said. “The number one thing that came to mind to us is the removal of the parts from the machine. What we’re talking about is trying to have this flexibility where our robots can move and have other functions, to give them flexibility of tasks.”
Developing the culture
Like any change in the workplace, Karn said it’s paramount that everyone buys in, and for management to ease concerns about the new technology for those employees who feel threatened by it.
“All the experience can be taught,” Myrand said. “It’s the will and the desire and the drive to do it. You need upper management support, which means money.”
There are many online training options that companies can utilize. And McGrath added that companies have simplified the data gathering process into easy-to-use tools.
“A lot of people when they think about Industry 4.0 and manufacturing analytics in general get scared and think it’s heavy coding and things like that,” McGrath said. “Most analytics companies have converted the hard coding portions of it to a drag-and-drop kind of environment. As complex as these data elements are, the tools and techniques that are out there today allow for companies to train their work forces to operationalize the types of insights needed.”
Bockman said it’s important to listen and utilize the experience already in the company. She found that it’s easier to teach the software to someone already rooted with manufacturing experience than the other way around.
“You want to utilize the experts, the people within your facility, because a lot of what you’re going to be measuring, they already know what they’re looking for,” Karn said. “It’s just a matter of how you get it and extract it. If you look at a process tech on an injection molding machine, he understands exactly what’s influencing this part with a short shot. They need to be involved to help identify what these key things are.”
Multiple attendees questioned the security of these programs, and overall the panel said those concerns could be mitigated by taking some basic steps.
“Certainly cybersecurity is on the mind of a lot of folks,” McGrath said.
Bockman said hacking is typically not likely, but there are steps companies can take to prevent it. For instance, warn employees not to use the equipment to charge their phones as that could open it up to possible vulnerabilities.
“The first thing to do is to train your staff on what not to do and what to do, very simple things,” Bockman said. “What we would do is go in and do an audit of the facility and put a report together on all of its vulnerabilities. There were harder things, especially in securing the older equipment, but there are industrial firewalls you can put on your system.”
It varies as to when a company will see a return on investment with automation and data harvesting, but most of the panel agrees that it usually occurs within a year. One thing Myrand warned was that automation is not a replacement for lean manufacturing processes, advising companies to first fix the problem before automating it.
“It obviously depends on the investment and how far you’re going with this,” Karn said. “If it’s pure data collection, just pulling things out of the things you have doesn’t have a lot of cost involved with it. When you start investing in additional software and newer equipment that’s adaptive, we’ve found usually within less than a year we see a payback.”
Myrand said at the end of the day, it’s all about staying competitive in the marketplace.
“Indirectly, they’re forcing it by driving for a better and better piece price,” Myrand said. “We have to constantly be optimizing our costs, so lower labor, less scrap and high productivity of our machines to keep up with the piece price demands they give us.”
Copyright 2019 Crain Communications All Rights Reserved.
Quality News Today is an ASQ member benefit offering quality related news
from around the world every business day.