Lean manufacturing principles and automation systems can coexist, although many lean purists contend that lean goals conflict with using automation. Smart applications of automation, however, can result in deployment of systems that are both automated and lean, with flexible manufacturing systems that can be easily reconfigured as factory operations change.
Flexible automation solutions are key to manufacturers incorporating automation geared for a lean manufacturing environment. Proper testing is essential when evaluating potential lean automation systems, according to Jamie Flinchbaugh, a founder and partner of the Lean Learning Center (Novi, MI), who advises employing a scientific approach to analyzing manufacturing process needs.
Manufacturers attempting to apply lean principles to automated systems should perform Plan, Do, Check, Act (PDCA) steps, the process of developing and testing a hypothesis, prior to fully implementing lean automation, notes Flinchbaugh, a lean practitioner and co-author of the book The Hitchhiker’s Guide to Lean. “Everyone, as they work on improvements, whether it’s a business process or it’s a manufacturing process, they’ll figure out where technology helps them,” Flinchbaugh says. “My big caveat is that you really need to test the solution.
“It’s one thing to say ‘Here’s technology that should solve this problem.’ It’s another thing to know that it solves it, and doesn’t make the situation worse,” Flinchbaugh adds. “The same effort of observation—directly observing work and understanding the real problem statement, and doing improvements with experimentation—perhaps even applies more here. It’s very easy to say: ‘This technology is designed to solve this problem,’ and all it does is make the problem worse and add three more problems to it.
“It’s that sort of experimentation in the Plan, Do, Check, Act cycle that I think we most often miss when we using technology,” Flinchbaugh states. “We also don’t test the concept, we don’t test the idea, before we jump to the solution, and so we really have to get better at finding cheap and easy ways to test the idea. Let’s just say it’s computer technology—if I can’t make the process work on paper, all the computer does is make it go faster. So if it’s a broken process on paper, it’s just going to be a broken process going faster on the computer.”
While much debate continues about automation and lean, manufacturers should put the myths to rest and instead focus on solutions, says Flinchbaugh, who recently discussed the topic of teaming automation technology with lean practices in a podcast interview featured on the web site of automation developer Bosch Rexroth Corp. (for more information, see www.boschrexroth-us.com/lean). “My experience is that implementing lean in a lot of manufacturing environments means that it really is an approach to solving manufacturing problems,” he notes, “and if the best solution to a particular problem involves automation, then that makes automation lean.”
In the lean toolbox, error-proofing, or poka-yoke, is a key element of lean manufacturing’s continuous improvement efforts. In addition to deploying PDCA, manufacturers must look for flexible automation systems that are easily modified and that also improve the process, save time, and reduce errors, says Flinchbaugh. “First and foremost, we really want that technology to be inherently flexible. Lean is about improvement, and if it’s not flexible, we can’t improve.”
Improving manufacturing quality is about reducing errors and defects, he adds, and error-proofing is a great tool that can be used, leveraging automation. “A great example is vision systems; they can detect when something is mounted incorrectly, or conveyors or carriers that can only receive a part in the exact orientation needed to have the part processed in the next step correctly,” he observes. “It absolutely should be part of our toolbox of solutions for making improvements in lean. By using automation, we prevent quality defects from occurring in the first place, or once they do, stop them from continuing. There are many things that automation can do, that other solutions can’t do, and automation can do them well.”
Lean practitioners’ skepticism of automation stems partly from the idea that lean systems generally involve manual work processes, and automation doesn’t fit into that equation. “I guess it depends a little bit on the kind of automation that you’re talking about, because many times people think of automated conveying systems as being the most wasteful part of a hard-automated system,” notes Kevin Gingerich, of the Bosch Rexroth Linear Motion and Assembly Technologies unit (Buchanan, MI) of Bosch Rexroth Corp. (Hoffman Estates, IL). “In many cases, that kind of thinking results from an outdated understanding of how conveyors are made and deployed, and the types of conveyors that are available.
“In the lean argument that manual processes are best, the thinking goes that inventory is waste and, therefore, work-in-process inventory is also waste,” Gingerich explains. “So if something is on a conveyor system being transported from place to place, it’s simply automation of the waste.”
With lean automation systems, Bosch Rexroth contends that lean production is about more than simple manual assembly cells, and is actually about the elimination of wasted time, wasted motion, wasted production. While the company offers a full line of automation systems, Gingerich says the decision on whether to automate a lean manufacturing operation depends on the situation.
“We don’t believe that every process should be automated just because we’re people who sell automation systems,” he adds. “We believe that there is a best process that you can find through testing, discovery, analysis, and trying. In some cases, you really don’t want to automate a process, because you can do it effectively with the manual labor that you have if you just rearrange the work and balance the work, and try to minimize travel distances, then simply use lean tools to create a more efficient assembly system or production system. There are some products, however, that may require gentle handling that a human just can’t carry off.”
Certain industries like medical or semiconductor manufacturing fit that scenario, where manual labor can’t duplicate what an automated cell can do. “That would be an example,” adds Gingerich, “or let’s say you have some kind of a gluing process that has to happen, and you’re trying to apply a thin bead of adhesive along a very narrow edge of some component, to mate the two components together. How good is a human going to be at guiding a glue gun along this top edge without slopping the glue over?
“So you may want to have a robot even in a fixed automation cell that does only that process—to make sure the process gets done quickly and accurately. There are a lot of lean thinkers who are probably re-establishing a connection to lean in situations like that. When it comes to conveying systems, there are many people, many lean gurus and advisors who think of conveyors when they need to design a synchronous system—the speed of the work is tied to the speed of the conveyor belt.”
Proper analysis of processes must precede any addition of automation to lean manufacturing operations, he cautions. “Your first tool, if you’re going to explore lean, probably should be a stopwatch, so that you can just watch someone do what it is you need to do,” Gingerich states. “You just time each piece of a manufacturing process or an assembly process, and you start to understand the wasted time involved.
“People are getting much smarter about using lean techniques in the manufacturing world. But the world doesn’t just shift on a dime; you have a lot of people who have been trained to do manufacturing one way, and they’ve been doing it that way for five or 10 years, and often you can meet up with resistance when you’re starting to re-work the processes and install new equipment, or change the way things happen.”
Resistance to change happens often in a manual environment, adds Gingerich, as manufacturing employees remain comfortable with their work processes. “A mistake that companies might make is to is just automate that process, because they’re sick and tired of dealing with intransigent workers or backsliding or something like that, when really the company needs to find the best method—they need to implement that method, and that might be a manual method or it might be an automated method—only the situation is going to determine that. If it does turn out that automation is the best way, it’s important to look for a flexible system that isn’t welded steel; maybe it’s bolted-together technology that lets you change it around, if you need to change your process later, because you’re probably going to. That’s what lean is all about, continuous improvement.”
Smart automation (autonomation) is one of the pillars of lean systems, along with waste reduction and Just-In-Time manufacturing. By deploying smart, flexible automation in manufacturing processes, manufacturers can help improve product flow and quality while improving factory equipment uptime.
Lean manufacturing systems are complemented by the application of robots in many ways, according to Dick Johnson, general manager, material handling, Fanuc Robotics America Inc. (Rochester Hills, MI). Newer vision systems add intelligence to robotics and play a big role in increasing the quality of parts, which can optimize lean manufacturing.
Using robots in machine-loading applications helps reduce waste and boost machine uptime, Johnson notes. “Machine utilization is enhanced, because the robot loader is faster than the operator. It’s faster since you don’t have to shut off the coolant, you don’t have to open the door, you can actually mount the robot right inside; the robot’s going to take the raw part in, and when it grabs the finished part, it’s going to do a quick rotate and the machine is back to cutting a part. I guarantee the robot will always, always beat the operator.
“A second point is that quality is going to improve,” Johnson adds. “In lean, one of the seven wastes has to do with defects. That’s going to improve for a couple of reasons; because the robot continuously loads the machine, as long as there are parts, the machine comes up to temperature and is happy—it makes better parts. It makes worse parts at the beginning of production. As the robot loads a machine tool, it doesn’t take a morning break, a lunch break, or the afternoon break—the machine is running continuously, it’s happier, and you’re making better quality components.”
Vision systems are now deployed on nearly a quarter of Fanuc’s material-handling applications, which is a big boon to quality improvement, notes Johnson. “With vision, the robot can check its own quality. The use of vision is basically exploding, and if you’re going after Six Sigma and you’re looking for that 3.4 visual defects per million, I guarantee the vision system’s going to find it.”
Lean and automation systems can exist together, Johnson states. “I had a visit from a large automotive component manufacturer that said they were doing lean. I said, ‘What’s lean mean to you?’ And they said, ‘We’re not going to use any robots!’ They were of the impression that a person was more flexible, and could more easily adapt to the variability in systems. I believe they were misguided. Basically, if you’re going to process something, you only have three choices: you can do it manually, and that can be a relatively simple application to a complex craftsman; you can do it with hard automation, which is relatively inflexible and it’s probably not going to help you with lean; or you can do it with flexible automation and robots. The robot can make the system leaner than fixed automation and help remove the operator from routine and repetitive tasks.”
Intelligent robots equipped with the latest vision offer highly flexible automation in lean environments, he adds. “You must give machines enough intelligence so when they are working abnormally, they flag the operation for humans. When this is case, humans would not have to monitor normal production, and would only focus on abnormal or fault conditions,” Johnson says. “Robots can accommodate one part or entire part families, as well as adapt to variation in part placement. That’s not the case in hard automation, where if you have a cocked part, you’re down, the red light comes on, and the job setter comes over, fixes the part that goes cocked twice an hour, and it’s back up and running. So the robot’s ability to adapt versus fixed automation can increase throughput in a lean system.
“Vision’s the key. You know, in days of old you had to fixture the parts for the robots so they’d be absolutely located, but today they just put it on the belt conveyor and it’ll find out where the part is and then pick it. If you want to do a part family, the vision system can first identify if it’s part A, part B, or part C, then it can identify the location of the part and instruct the robot to pick it. The robot can then tell the upstream equipment that this is part A, part B, or part C. Or it can use it to load a different machine or operation. Intelligence is the key. The big difference is that vision has become more intuitive and less expensive. I still remember when vision was $80,000, and then you needed about $80,000 more to do the application engineering. Today, vision is much lower cost. Our vision guidance is $5995, and our visual error-proofing is $2995. It’s also easier to program, so the expensive application engineering is no longer required.”
Vision capability is now been built into all Fanuc robots, and the company recently added a new Diagnostic Video Monitor (DVM) to help pinpoint errors that occur during untended operations, or when an operator is away from the cell during a break. The Fanuc DVM uses an iRVision camera attached to a Fanuc R-30iA robot controller. Users set up an event-based trigger using the teach pendant, according to Fanuc, and the system can capture digital images of a manufacturer’s parts or process for compliance and traceability requirements, using the captured video before and after events for root-cause problem-solving. “If you’re truly going to do lean, you require automation,” Johnson says. “You require flexible automation, and robots have a significant part to play in that.”
Safety issues in robotic cells also has been an issue, and Fanuc has a new package, its Dual Check Safety (DCS) Speed and Position Check software, that helps safely shrink the size of robotic workcells. The Fanuc DCS safety-rated robot software allows the safety design of the robot system to use the robot for some of the safety functions. “It allows the robot to do some self-checking with redundant hardware that provides a significant benefit to customers,” Johnson says. “It allows you to safely reduce the size of the cell—wasted space is Muda. In some cases, the size reduction is very significant. It may even make certain applications possible that weren’t, under the old rules.”
Welding automation supplier SmartTCP (Farmington Hills, MI) offers customers robotic welding solutions for high-mix, low-volume production. Its solutions help construction builders reduce inventories and shrink factory floor space, while speeding production of high-quality parts, says Efi Lebel, SmartTCP founder and CEO.
“One of the problems if you have a high-production volume is inventory,” notes Lebel. “You produce just in time and you reduce inventories, because in a flexible manufacturing environment, you want to produce exactly the mix that is needed to meet market demands. SmartTCP specializes in high-mix, low-volume production, a solution that can help you reduce inventories and build parts faster and more accurately. Another aspect of lean is that when you compare SmartTCP, it’s more efficient compared to manual welding—if you replace manual welding with automation, you absolutely gain benefits by reducing space and you save on consumables.”
New programming software from robotics developer Adept Technology Inc. (Livermore, CA) offers manufacturers a way to enhance optimization of automation and production lines in a lean environment. The Adept package, ACE PackXpert, is built upon its Automation Control Environment (ACE) programming environment, allowing users to easily add robots and other automation into lean systems, according to Travis Armstrong, Adept system engineer. The system is aimed at packaging applications, but Adept is planning to add similar systems for the solar energy industry and for general-assembly applications.
“As an automation and motion-control company, we often get left out in discussions of lean manufacturing, and I’ve always been a little bewildered at that,” notes Armstrong. “What we have always worked on is efficiency improvements. Automation is often seen as a tool to use when trying to go leaner, reduce waste, and optimize the production process. The software that we’re working on now is a continuation of that.”
Adept’s ACE offers users a framework to connect its products with other machines and systems on production line, notes Armstrong, a member of the SME Automated Manufacturing & Assembly Technical Community. “By connecting something that’s doing not only robotic motion and vision, but also communicating with your operator interfaces—whether they’re machine panels, HMIs, OPC, PLC, and machine tools, keeping track of conveyor belt systems, feeders, all in a single environment—it really makes things easier to do on a production line when you debug something or collect data,” Armstrong says. “This environment gives you the tools to connect everything—log data, communicate—even share this out to a production line management system, where the plant manager may want to see what every machine is doing on a line.”
This article was first published in the February 2009 edition of Manufacturing Engineering magazine.
Connect With Us
