Podcast: Manufacturer harnesses edge intelligence to build 'a core for the future'
[Editor's note: This podcast was originally published on Feb. 8, 2019.]
In manufacturing, slight dips in machine cycle times can mount quickly, affecting output and the bottom line. To optimize its production processes, Canadian manufacturer CuBE Packaging Solutions embarked on a transformation journey driven by IoT and edge intelligence that generates minute-to-minute insights on preventive and predictive machine maintenance, quality control, operations efficiencies, and more.
Key to the effort is a new infrastructure core that created a consolidated system of previously isolated machines and a platform that enables the company to scale and shift as market and technology changes dictate.
“It’s not just built for today's needsâ it's built for expansion capabilities. It's built for year two, year three,” says Len Chopin, president of CuBE Packaging.
Listen to this Hewlett Packard Enterprise Voice of the Customer podcast to learn how the manufacturer is using the Intelligent Edge and modernized systems to achieve continuous, factory-wide improvements.
Dana Gardner: Hello, and welcome to the next edition of the BriefingsDirect Voice of the Customer podcast series. I’m Dana Gardner, principal analyst at Interarbor Solutions, your host and moderator for this ongoing discussion on digital transformation success stories.
Our next manufacturing modernization and optimization case study centers on how a Canadian maker of containers leverages the Internet of Things (IoT) to create a positive cycle of insights and applied learning.
We will now hear how CuBE Packaging Solutions Inc. in Aurora, Ontario, has deployed edge intelligence to make 21 formerly isolated machines act as a single, coordinated system as it churns out millions of reusable package units per month.
Stay with us as we explore how harnessing edge data with more centralized real-time analytics integration cooks up the winning formula for an ongoing journey of efficiency, quality control, and end-to-end factory improvement.
Here to describe the challenges and solutions for injecting IoT into a plastic container’s production advancement success journey is Len Chopin, president at CuBE Packaging Solutions. Welcome, Len.
Len Chopin: Hi, thanks for having me.
Gardner: Len, what are the top trends and requirements driving the need for you to bring more insights into your production process?
Chopin: The very competitive nature of injection molding requires us to stay ahead of the competition and utilize the Intelligent Edge to stay abreast of that competition. By tapping into and optimizing the equipment, we gain on downtime efficiencies, improved throughput, and all the things that help drive more to the bottom line.
Len Chopin, CuBE Packaging Solutions
Gardner: And this is a win-win because you’re not only improving quality, but you're able to improve the volume of output. So it’s sort of the double benefit of better and bigger.
Chopin: Correct. Driving volume is key. When we are running, we are making money and we are profitable. By optimizing that production, we are even that much more profitable. And by using analytics and protocols for preventive and predictive maintenance, the IoT solutions drive an increase in the uptime on the equipment.
Tapping intelligence at the edge
Gardner: Why are sensors in of themselves not enough to attain intelligence at the edge?
Chopin: The sensors are reactive. They give you information. It’s good information, but leaving it up to the people to interpret [the sensor data] takes time. Utilizing analytics, by pooling the data, and looking for trends means IoT is pushing to us what we need to know and when.
Otherwise, we tend to look at a lot of information that’s not useful. Utilizing the intelligent edge means it's pushing to us the information we need, when we need it, so we can react appropriately with the right resources.
Gardner: In order to understand the benefits of when you do this well, let's talk about the state at CuBE Packaging when you didn't have sensors. You weren't processing, and you weren't creating a cycle of improvement?
Chopin: That was firefighting mode. You really have no idea of what's running, how it’s running—is it trending down, is it fast enough, and is it about to go down? It equates to flying blind, with blinders on. It’s really hard in a manufacturing environment to run a business that way. A lot of people do it, and it’s affordable—but not very economical. It really doesn’t drive more value to the bottom line.
Gardner: What have been the biggest challenges in moving beyond that previous “firefighting” state to implementing a full IoT capability?
Chopin: The dynamic within our area in Canada is resources. There is lot of technology out there. We rise to the top by learning all about what we can do at the edge, how we can best apply it, and how we can implement that into a meaningful roadmap with the right resources and technical skills of our IT staff.
It’s a new venture for us, so it's definitely been a journey. It is challenging. Getting that roadmap and then sticking to the roadmap is challenging, but as we go through the journey, we learn the more relevant things. It's been a dynamic roadmap, which it has to be as the technology evolves and we delve into the world of IoT, which is quite fascinating for us.
Gardner: What would you say has been the hardest nut to crack? Is it the people, the process, or the technology? Which has been the most challenging?
Trust the IoT process
Chopin: I think the process, the execution. But we found that once you deploy IoT, and you begin collecting data and embarking on analytics, then the creative juices become engaged with a lot of the people who previously were disinterested in the whole process.
But then they help steer the ship, and some will change the direction slightly or identify a need that we previously didn't know about—a more valuable way than the path we were on. So people are definitely part of the solution, not part of the problem. For us, it’s about executing to their new expectations and applying the information and technology to find solutions to their specific problems.
We have had really good buy-in with the people, and it’s just become about layering on the technical resources to help them execute their vision.
Gardner: You have referred to becoming the "Google of manufacturing.” What do you mean by that, and how has Hewlett Packard Enterprise supported you in gaining that capability and intelligence?
People are definitely part of the solution, not part of the problem. For us, it's about executing to their new expectations and applying information and technology to find solutions to specific problems.
Chopin: “The Google of manufacturing” was first coined by our owner, JR. It’s his vision, so it’s my job to bring it to fruition. The concept is that there’s a lot of cool stuff out there, and we see that IoT is really fascinating.
My job is to take that technology and turn it into an investment with a return on investment from execution. How is that all going to help the business? The Google of manufacturing is about growth for us, by using any technology that we see fit and having the leadership to be open to those new ideas and concepts. Even without having a clear vision of the roadmap, it means focusing on the end results. It’s been a unique situation. So far, it’s been very good for us.
Built for the future
Gardner: How has HPE helped in your ability to exploit technologies both at the edge and at the data center core?
Chopin: We just embarked on a large equipment expansion [with HPE], which is doubling our throughput. Our IT backbone, our core, was just like our previous equipment—very old, antiquated, and not cutting edge at all. It was a burden as opposed to an asset.
Part of moving to IoT was putting in a solid platform, which HPE has provided. We work with our integrator and a project team that mapped out our core for the future. It’s not just built for today's needs—it's built for expansion capabilities. It's built for year two, year three. Even if we’re not fully utilizing it today, it has been built for the future.
HPE has more things coming down the pipeline that are built on and integrated to this core, so that there are no real limitations to the system. No longer will we have to scrap an old system and put a new one in. It’s now scalable, which we think of as the platform for becoming the Google of manufacturing and which is going to be critical for us.
Gardner: Future-proofing infrastructure is always one of my top requirements. All right, please tell us about CuBE Packaging—your organization’s size, what you're doing, and what [your] end products are.
Chopin: We have a 170,000-square-foot facility, with about 120 employees producing injection-molded plastic containers for the food service industry, for home-meal replacement and takeout markets, distributed to Canada as well as the U.S., which is obviously a huge and diverse market.
We also have a focus on sustainability. Our products are reusable and recyclable. They are a premier product that come with a premier price. They are also customizable and brandable, which has been a key to CuBE’s success. We partner with restaurants, with sophisticated customers, who see a value in the specific branding and of having a robust packaging solution.
'A plant within a plant'
Gardner: Len, you mentioned that you're in a competitive industry and that margin is therefore under pressure. For you to improve your bottom line, how do you account for more productivity? How are you turning what we have described in terms of an IoT and data capability into that economic improvement to your business outcome?
Chopin: I refer to this as having a plant within a plant. There is always a lot more you can squeeze out of an operation by knowing what it’s up to, not day by day, but minute by minute. Our process is run quite quickly, and so slippage in machine cycle times can occur rapidly. We must grasp the small slippages, or predict failures, or when something is out of technical specifications from the injection molding standpoint or we could be producing a poor quality product.
Getting a handle on what the machines are doing, minute by minute by minute, gives us the advantage to utilize the assets and the people and so to optimize the uptime, as well as improve our quality, to get more of the best product to the market. So it really does drive value right to the bottom line.
Gardner: A big buzzword in the industry now is artificial intelligence. We are seeing lots of companies dabble in that. But you need to put in place certain things before you can take advantage of those capabilities that not only react but have the intelligence to prescribe new processes for doing things even more efficiently.
Chopin: We are already embarking on using sensors for things that were seemingly unrelated. For example, we are picking up data points off of peripheral equipment that feed into the molding process. This provides us a better handle on those inputs to the process, inputs to the actual equipment, rather than focusing on the throughput and of how many parts we get in a given day.
For us, the AI is about that equipment uptime and of preventing any of it going down. By utilizing the inputs to the machines, it can notify us in advance, when we need to be notified.
Rather than monitoring equipment performance manually with a clipboard and a pen, we can check on run conditions or temperatures of some equipment up on the roof that's critical to the operation. The AI will hopefully alert us to things that we don't know about or don't see because it could be at the far end of the operations. Yet there is a codependency on a lot of that pre-upstream equipment that feeds to the downstream equipment.
So, for us to gain transparency into that end-to-end process and having intelligence built in enough to say, “Hey, you have a problem—not yet, but you're going to have a problem,” allows us to react before the problem occurs and causes a production upset.
Being more intelligent about gathering intelligence
Gardner: You can attain a total data picture across your entire product lifecycle and your entire production facility. Having that allows you to really leverage AI.
Sounds like that means a lot of data over a long period of time. Is there anything about what's happening at that data center core, around storage, that makes it more attractive to do this sooner than later?
Chopin: As I mentioned previously, there are a lot of data points coming off the machines. The bulk of it is useless, other than from an historical standpoint. So, by utilizing that data—not pushing forward what we don't need, but just taking the relevant points—we piggyback on the programmable logic controllers to just gather the data that we need. Then we further streamline that data to give us what we're looking for within that process.
It's like pushing out only the needle from the haystack, as opposed to pushing the whole haystack forward. That’s the analogy we use.
Gardner: So being more intelligent about how you gather intelligence?
Chopin: Absolutely! Yes.
Gardner: I’m afraid we’ll have to leave it there. We have been exploring how a Canadian maker of containers is leveraging IoT to create a positive cycle of insights and applied learning. And we've learned how harnessing edge data in centralized, real-time analysis can cook up a winning formula for an ongoing journey of efficiency, quality control, and ongoing, end-to-end factory improvement.
Please join me in thanking our guest, Len Chopin, president at CuBE Packaging Solutions in Aurora, Ontario. Thank you so much, Len.
Chopin: Thanks for having me.
Gardner: And a big thank you as well to our audience for joining this BriefingsDirect Voice of the Customer digital transformation success story. I’m Dana Gardner, principal analyst at Interarbor Solutions, your host for this ongoing series of Hewlett Packard Enterprise-sponsored interviews.
Thanks again for listening. Please pass this along to your IT community, and do come back next time.
IoT, data, and AI boost manufacturing uptime: Lessons for leaders
- Minute-by-minute visibility is crucial to improving manufacturing productivity.
- Slippage in machine cycle times can mount rapidly; detecting small slippages is critical.
- Data points from peripheral equipment can complement data on throughput to reduce reactivity.
This article/content was written by the individual writer identified and does not necessarily reflect the view of Hewlett Packard Enterprise Company.