Podcast: How to identify and eliminate production bottlenecks

It’s a story so familiar that it is also a children’s proverb: “for want of a nail, the kingdom was lost.” In this new podcast episode we cover how to identify and eliminate pinch points on the manufacturing floor. Our guests this week, Jerry Nation and Hunter Price of Ingenics Corporation, talk about their experience helping plants in the automotive sector get past production bottlenecks that were dramatically slowing down production, and stressing out downstream operations teams.

Below is the transcript of the podcast:

PS: You guys have a pretty broad view of the way industry operates, so maybe we can start the podcast by each of you telling us a little bit about yourselves, and the kinds of specifically asset management or maintenance work that you've done with clients via Ingenics.

JN: My name is Jerry Nation, I'm a key account manager with Ingenics. I have about 15 years experience in the industry. I've worked with quite a few automotive and aerospace OEMs over this years. We generally specialize in efficiency improvements, equipment installations, logistics, supply chain management, pretty much just anything across the board with these guys. We tend to follow industry trends and really focus in on what the pain points are for our customers. I've worked personally in all those different fields, everything from supply chain management to resource allocation to equipment installations, preventive maintenance work. You name it, we've touched on it.

HP: Thanks, Jerry. My name is Hunter Price, I'm the assistant key account manager here in Charleston as well. And yeah, I've been in the industry and within Ingenics for about 5 years now, same as Jerry, worked various different roles with various different customers in the automotive industry. My expertise is spanned from logistics to supply chain planning, warehouse planning. Data analytics is a big part of it as well, and data transformations for our customers. As of now, seems to be maintenance and preventative maintenance is a big key point for a lot of our customers, so happy to chat about it more.

PS: That sounds great. I know a lot of people are going to be listening in. I'm curious to hear what you're hearing about other plants in the sector. You know, individual teams don't often have visibility into either other verticals or even other plants in their vertical unless they go to events and get to chitchat with people who are there to share experiences.

Before we get into specific details of some of your engagements, let's talk about those trends in industry. What are some of the pain points that you're seeing now since we're living in a post COVID (post COVID crisis, I should say) age where you've even got supply chains starting to straighten themselves out. I'm curious what you're hearing, what your clients’ key pain points are in this regard.

JN: A lot of times what we see is those supply chains do start to straighten out, and they're trying to ramp the production numbers back up to what they were before COVID. There's a lot of pressure there, right. These lines, they're not used to running at those volumes anymore. Everyone kind of slowed down, and they got used to the smaller volume. Now they're trying to push the gas pedal, they're trying to get back to where they were, and number of units at the end of the line is really what matters.

So what that does for our maintenance guys is it really pushes their piece to the back burner. They’re being encouraged to really just get the volume out, not to do the root cause analysis, not to do the Five Whys, not to focus on preventive maintenance. What we see happen there is the equipment, especially in a robotic setting where you have a lot of complexity, whenever they have a downtime issue where they stop, they're just clearing the code as quickly as they can trying to get the machinery running again without doing that root cause analysis, and then these issues, they start compounding. What could have been a relatively simple fix on the line becomes bigger and bigger and bigger as this compounds to different robotic cells, different stations throughout the facilities, and it gets to be a topic that's so big that our customers don't really know how to get ahead of it any more.

PS: That's fascinating because this reminds me of a presentation that I heard 10 years ago with the manufacturer. It was those little K cups for coffee, I'll keep the brand name private, but they were asked to get 30 lines ramped up within the space of about a year and then retrofit the maintenance program. So I'm curious to know that you're seeing this not just with one or two plants, but this is a general problem people are having across the board, to figure out how to not let maintenance processes pile up too badly, right?

JN: Yeah. And it's only compounded by an industry that's starting to look at more economical ways and greener manufacturing processes, especially as we start shifting to electric vehicles. That adds a whole another level of complexity, so a lot of the established OEMs, they've been doing things certain ways for many, many years. In addition to the COVID impact and the supply chain impacts and this new ramp up, you also have complexity changes – you're dealing with high voltage components, you're dealing with stuff that that no one's ever had to deal with and a lot of times the regulations on that stuff are still being sorted out. They might not understand on the maintenance side exactly what PM work has to be done on this new carrier because it's something different from what they're used to in the past, because it's handling different components, different weights, different types of situations that just compound these problems we're seeing.

PS: Maybe we can turn to a specific example, and we'll keep the name of the plant and the company private. But you had you have an automotive client in the Charleston, SC area that we talked about a little bit on a pre-call, and you both have mentioned that a significant portion of your work at that plant was centered on helping the body shop out, both when it came to data management and also just increasing production. Hunter, since you work more on the data analysis side, could you tell us a little bit about what some of the data challenges were that you tackled with this customer?

HP: Definitely. One of the big things with body shops, and it really again spans across all of our customers that we've seen, is the volume of data, the integration of data, and the variety of data that come in from a body shop. So imagine an automotive body shop with dozens to hundreds of various robots, sometimes of the same manufacturer, sometimes of varying manufacturers, and then you have different PLCs and different coding languages in there.

There's so many different tunnels for data to come into for a manufacturing facility, and we see a lot of them struggle with integrating them all into one platform. They start buying one robot for several years, and then they start buying another robot for several years, and now they have two different systems feeding their data. One of the challenges we face is finding a way of getting all the data into the cloud, finding ways that we can help integrate our customer’s data or just collect it in the best way possible. Again, for people that don't know, these body shops and these assembly lines, they're running pretty much 24/7 and you're getting data every second about your robot’s performance, about the quality of the vehicle, about the temperature in a work cell – so much different data.

Where we come in is helping them pick and choose what data is relevant, what's going to help them improve their bottlenecks or find their bottlenecks, or find the robots that are the biggest issue. Once you're able to collect all that data, then you can get into the real issue of predicting when failures are going to occur to your robot. There's a lot of preventative maintenance out there, a lot of monitoring, but what we see now, one of the trends is people predicting when a robot might break down so they can go and check it beforehand.

When you have all the data and you know your limits, your upper limits, your lower limits, and you know what's a good robot process versus a bad one, you can use machine learning nowadays to say, “hey, with this robot, 10 or 15 cars might be an issue,” and we're able to attach and attack those issues before they occur. Then you don't have these line downs, you don't have to wait an hour to figure out what the issue is, or where the issue is. Collecting the data is the biggest piece for us, helping our customers collect it, but then also helping them know what can you do with these 30 million rows (of data) you received in the last five days.

Another one of the challenges is with real-time data. You're going to want, especially in a body shop, real-time data so you're not seeing, hey, an hour ago, this one started to act a little funny. A tip for people setting up or preparing to increase their productivity is honestly making sure that the data, the back end of the data is set up properly. You don't want to have everything set up and then all of a sudden you're like, “where do we put all this data? We have 10 different systems collecting!” There's a lot of challenges with collecting and analyzing data.

PS: You mentioned something important, which is that with all these data going in various places and not being assembled in one location, it's really difficult to see which are the stand out data points. I think you both mentioned before that building dashboards with this client really helped out in a couple of ways. On the data side, it helped make the problems visible that much more quickly. Then Jerry, you mentioned that there was an issue where not everyone spoke the operational language of the body shop, so those dashboards helped communicate to everyone else in the company on a data level exactly what was happening there.

JN: Yeah, exactly. We see that a lot of times as well, where when there's an issue in one particular shop, then they try to pull some resources from different areas to help out. You want to leverage all the experience you have, get all the help you can when you have a critical situation. But sometimes some of those resources you bring in, they're not going to be experts in that particular shop because they're going to be coming from paint or assembly or somewhere else. Having a dashboard like this that's able to quickly and concisely convey that information can be a huge win for a company. Because then those new resources that added capacity, they're really able to help out a lot more because they can clearly see what the issues are without necessarily having to have a really detailed understanding of that that particular shop’s operations.

PS: With complex processes like automotive too, you've got shops downstream of, say, a body shop which are also impacted by delays in the body shop, and that was an issue with this plant as well, right? There were some shifts that were actually being impacted when not enough parts were moving from one team to the next.

JN: Oh yeah, that's definitely an issue we see at every customer across the country or around the world. If a body shop isn't producing, you can't get the vehicle painted, you can't get your bodies to the paint shop, then you can't get your painted bodies to assembly. So like you said, it all falls downhill.

PS: After you guys worked with this client, how did all that process end up resolving itself? Were you able to get the production ramped up, where say the paint shop team wasn't as impacted?

HP: It was specific to this customer here, but we've used it other places as well, is that with the data we're pulling in, they know they have an issue. They know there's an issue somewhere and we're pulling in millions of rows of data, but that’s all they get. So when we came in, were able to build one of these dashboards that showed over the last seven days every single minute where the issues were occurring. When we came in, it was just various millions of rows of data about robots and vehicles, and when we left there, they had a dashboard that showed exactly what stations at what times were causing the issues, and then they could use that to break that bottleneck and now they can see downstream. What's the next bottleneck? And they're able to continuously find their bottlenecks, and go in and break them without having to just guess.

JN: To elaborate on that a little bit, we like to be strategic partners with these customers. We don't really hit a point where we walk away from them and say, hey, this good enough. The issues I would say, they're continuing to get better, but we're still actively involved, we're still putting our resources in there, trying to get them not just to the level they were at, but to the next level – future proofing, getting ready for increases to the line speeds, to new model launches. We're always thinking 2, 3, 4 years down the road along with those customers. We see that improvement, we see them getting better, and we're pushing them to that next level beyond it.

We'll get you out here on this question, guys: what advice would you give to listeners who are currently working on maintenance and reliability process improvement initiatives? Is there something that stands out right now that you think people can do in response to current economy and supply chain issues that would be really useful to them?

JN: My biggest thing is, don't skip the root cause analysis. It can be really simple and easy when you have all this pressure to just clear an error code and try to move on to the next problem. But really, you've got to slow down a little bit because taking an extra few minutes today saves you hours tomorrow. That would be my biggest advice: slow down just a little, really pay attention to your root cause analysis, understand your issues, find your corrective action, get it implemented, and make sure you're following a robust process.

HP: Yeah, I couldn't agree more. I always would say, our customers and really us, we try to focus on that continuous improvement culture. Like Jerry said, we don't just go in and leave. Once you fix one issue, there's another one coming and we're there to always improve our customers. Having that mindset that it's never perfect is key in automotive, because there's always room for improvement somewhere, or somewhere to cut cost.

Other than that, having that proactive approach would be leveraging the technology like I talked about earlier, building up your data management system early so that when you do have all the data you've ran for a few months or a few years, you can analyze it and really use machine learning to the best of your abilities to catch issues before they arise.

About the Podcast
Great Question: A Manufacturing Podcast offers news and information for the people who make, store and move things and those who manage and maintain the facilities where that work gets done. Manufacturers from chemical producers to automakers to machine shops can listen for critical insights into the technologies, economic conditions and best practices that can influence how to best run facilities to reach operational excellence.