Think automation is too difficult in a high-mix, low-volume environment? Think again
Robots and job shops have not typically been talked about together. After all, everyone knows that automation is only suitable for high-volume production, and the typical mom-and-pop operation is anything but—its schedule filled with orders for high-mix, low-volume, and often highly complex work. Who needs a robot in this environment, where setups are performed several times a week, or even each day?
Before you skip to the next article, though, you might want to consider a few things. For starters, many experts suggest that, regardless of what the owners might tell themselves, job shops suffer the lowest machine utilization in the industry. Overall equipment effectiveness (OEE) values of 50 percent or lower are not unusual. Automation is a clear solution to bettering this number.
And because machine changeover is more frequent in job shops, more skilled machinists and programmers are needed to perform setup-related activities—people who are commanding higher salaries and are difficult to find. This drives up operating costs, reduces opportunities for additional work, and makes around-the-clock production extremely challenging. Given the shortage of skilled workers, automation should, therefore, be the Holy Grail for any job shop, even if implementation might not be an easy path.
Possibilities for Plug and Play
The question is, where do you begin? Zach Spencer has a few ideas. A custom proposal engineer for Methods Machine Tools Inc., Sudbury, Mass., he noted that automation can be close to a plug-and-play affair, especially when included as part of a new machine tool purchase.
“We developed our JobShop Cell nearly 15 years ago, and it’s since become one of our most popular automation solutions for the FANUC RoboDrill,” he said. “It’s designed for shops that have at least some changeover every day, are restricted on floor space, and need the flexibility to switch easily from one part to another.”
The JobShop Cell is available in two configurations, he explained. One uses an over/under conveyor arrangement to bring raw material to the robot and take finished parts away, while the other uses a cabinet containing a series of drawers for workpiece and material storage. Both employ a FANUC robot to load and unload the machine (or pair of machines, in the case of Methods’ Twin RoboDrill JobShop Cell), and both provide a way to leave a CNC unattended for hours or even a complete shift, depending on the workpiece.
For higher volumes, Methods offers its Plus-E, a multi-pallet, elevator-style automation system for mid- to high-volume jobs, as well as the Plus-K for shops that wish to automate machine setup as well as operation, with the ability to swap out toolholders, pre-loaded vises, and fixtures. “You can tell the system that these five vises are Part A, the next two vises are Part B, the one after that is Part C, and the next ten will be Part D,” Spencer said. “After that, you just load it up with material and go home. Both of the Plus Series machines are self-contained units that sit alongside a RoboDrill, and are perfect for shops looking for an easy way to step into automation.”
Take The Trolley
Methods isn’t the only provider taking this approach. Eric Nekich, responsible for operations, technology, and inside sales at Lang Technovation Co., Hartland, Wis., said the company’s RoboTrex automation systems use a FANUC robot and up to four movable trolleys, each containing a number of pre-loaded vises. The RoboTrex 52, for example, can store up to 42 vises per trolley (based on part size and trolley configuration), potentially offering unattended machining of 168 workpieces.
Like the JobShop Cell, the RoboTrex is self-contained and fully guarded, but can also be integrated with a variety of machining center brands and models. The only hurdle is wrapping one’s head around the need to invest in a few of Lang’s Makro-Grip five-axis vises, required for use with the system (and made, coincidentally, on 50 RoboTrex-equipped machining centers at Lang’s factory in Germany). “It’s an investment with a very quick ROI, however, because your spindle uptime will drastically increase with the RoboTrex USA system,” said Nekich.
Michael Gamache is the business development manager at New Berlin, Wis.-based Gamache Systems, a division of Metalcraft of Mayville and the exclusive integrator for RoboTrex systems in the United States. He recognizes that some customers might balk at investing in a new workholding system, but offered some excellent reasons to do just that.
“As a business owner, I understand the fact that vises aren’t inexpensive, but from an automation perspective, high-quality workholding is half the battle,” he said. “You need to get the raw material to the robot in a repetitive, accurate manner, and the Makro-Grip system is one of the best ways to accomplish this. Furthermore, the system is very easy to set up and operate because the robot is handling the vises and not a wide variety of raw material sizes; because of this, one program works for every part.”
Get a Grip
Tyler Mardaus, a design engineer at Minneapolis-based Kurt Manufacturing – Industrial Products Division, agreed on the need for high-quality workholding, even though his company’s approach is significantly different than that proposed by Lang. “Our Custom Engineered Workholding (CEW) team designs a lot of solenoid-controlled, double-acting hydraulic systems,” he said. “Hydraulics provide higher psi and therefore greater clamping force than manual workholding, while the solenoid supports electronic activation of the vise through the robot or machine control, creating a truly automated system.”
He added that shops should start their automation journey by targeting the highest-volume jobs. And if there are parts that can use a vise (or multiple vises) with machinable jaws, these can be swapped out during changeover, reducing setup time. As for investment, there’s no avoiding the need for a hydraulic pump and plumbing, but it might be possible to retrofit a shop’s existing manual vises with hydraulic-friendly nut-and-screw assemblies.
“You should strive to create a closed-loop control system,” Mardaus said. “This begins by installing some sort of part-seating confirmation device, but you also need to verify that the vise has been clamped at the correct force before allowing the cycle to start, and that the vise is unclamped before the robot attempts to remove the workpiece. All of these steps are crucial to any automated machining solution.”
Speeding Up Workholding Exchange
Mardaus’ comment on setup time reduction is spot on. After all, why bother reducing labor costs through automation if you’re going to incur hours of expensive downtime every time you change jobs? The quick-change jaws he mentioned are an excellent place to start, but if the budget permits, a zero-point positioning system, like that of Lang’s Makro-Grip (and others), turns workholding exchange into a one-minute exercise. There should also be robust CAM capabilities, with simulation software that evaluates the post-processed toolpaths, toolholders, cutting tools, workpieces, workholding, and the machine tool itself.
And as Dietmar Moll, director of business development at Zoller Inc., Ann Arbor, Mich., explained, shops should place a high priority on tool management, starting with an offline presetting system and associated software. “One of the biggest benefits of any automation strategy is that machine tending and other non-value-added work is taken over by a robot or cobot, freeing qualified personnel to do more important tasks,” said Moll.
Offline tool presetting is one of these tasks, but there’s also thorough job planning, toolholder maintenance and organization, cutting tool inspection, and analysis of shop floor data. To this end, Zoller offers a variety of solutions well beyond toolsetting. Its Bronze, Silver, and Gold level tool management solutions (TMS) allow a shop to increase its capabilities as needed, delivering significantly higher visibility to production processes and tooling-related assets. And because these can communicate with Zoller’s »cora« collaborative robot, Moll added, they serve to “close the loop” by automating toolroom tasks such as tool cleaning and storage.
Similarly, the company’s integrated tool storage solutions help to organize these assets, further reducing downtime and inefficiency. And when the job is complete, these same systems can be used to check tools back into the crib, review their performance, and record any remaining tool life. “This helps shops better understand which cutting tools are working well and which ones aren’t, giving them the ability to further improve processes and reduce tooling costs,” he said.
Robust tool management is also an essential part of job scheduling, something critical to machine tool and robot utilization alike. To Moll, this means knowing what jobs are next in line and what tools are needed to run them. To Daniel Carranco, it means more than that, especially in the context of automation. The director of continuous improvement at Global Shop Solutions Inc., The Woodlands, Texas, he noted that all shops—automated or not—should have a good ERP and scheduling system in place, one that can also monitor the shop floor and collect information on the machining process.
Achieving the monitoring and data collection part of that equation requires integration-friendly ERP software, machine tools, and robotics, though. “We’ve had situations where the customer wanted to integrate their equipment with our system but found that the software platform used by that equipment was almost impossible for us to work with,” he said. “So the first bit of advice for anyone pursuing an integrated production environment is to make sure that everything can communicate effectively, and that everyone is clear as to the kinds and volumes of data that will be exchanged.”
Communication, it seems, means more than it once did. Where ERP systems have long excelled at collecting data from the humans on the shop floor, they now must gather it from robots and machine controllers as well. Further, this data not only contains the ubiquitous job number, production time, and quantity completed (or scrapped) that users have come to expect, but might also display real-time material consumption, inspection and part quality information, machine and robot status, and so on.
“There’s certainly much more information available today than there once was, and given the right ERP software, equipment, and integration, all of it can be seamlessly and automatically collected,” he said. “Doing so provides completely new opportunities for continuous improvement, but more importantly, it frees employees to focus on more important things. There’s no more time wasted on manual data collection, which is often spotty anyway, and people can work on process planning and analysis.”
Giving Best Practices a Boost
As Carranco suggested, these technologies should be pursued regardless of a shop’s plans to automate, as all serve to make operations more efficient and its OEE levels much higher. And when the next step does get taken—implementing a robot or material handling system—the investment will pay off that much faster. Assuming you don’t care for or can’t use any of the pre-configured systems already discussed, the question then becomes, “what robot should we buy, and how should we implement it?”
Dean Elkins can recommend one. A segment leader for material handling at the Motoman Robotics Division of Yaskawa America Inc., Miamisburg, Ohio, Elkins said job shops often present a range of suitable robotic applications, whether it’s machine tending, part deburring, automated assembly, or robotic welding. Finding the right equipment for these applications, though, requires a good deal of evaluation and talking to knowledgeable people.
“Justifying an investment in automation depends on several factors,” said Elkins. “Obviously, the larger the production quantity, the more it makes sense to automate it. But even lower volume, repeat work can be profitable, provided the shop takes the necessary steps to reduce setup time as much as possible. Parts with longer run times are also good candidates, as are parts with simple handling requirements.”
Elkins noted that most job shops find a small to medium-sized floor-mounted robot—meaning a payload of 5-25 kg—is all that’s needed to deburr a batch of parts, weld a crateful of brackets, or tend a CNC lathe after everyone’s gone home for the night. With that, however, some level of safety guarding will be required, as will grippers, an interface to the machine tool, and possibly programming and simulation software.
And don’t forget the training, he warned. Automation is much simpler than it once was, yet the savvier job shop owner is still going to make an additional investment in training, and have a dedicated person on staff to manage the shop’s robotics. That said, collaborative and force-limiting robots are becoming quite popular, especially where floor space is a concern. Robots can be equipped with self-changing or quick-change grippers, increasing flexibility. Vision systems can often eliminate or at least simplify part conveying systems, and the ability to teach a robot its tasks helps reduce the need for offline programming systems. “There’s a lot to consider,” said Elkins. “That’s why I encourage anyone interested in automation to do their homework, ask questions, and partner with a reputable integrator.”
Acieta LLC, Waukesha, Wis., is one such company. Steve Alexander, vice president of operations, warns that not all integrators are created equal. “Make sure you work with someone who has experience with your type of work and the skill level to implement your robotic automation project,” he said. “Get references, go see their previous installations, and understand the quality of their work. Industrial robots are designed to last well over 10 years, so you want to make sure everything else in the cell is designed for an automated environment and is just as reliable as the robot itself.”
Doing so can lead to a return on investment (ROI) of just a few months, although a year or two is probably a more realistic expectation. Much of that depends on the application, of course, but the type of robot also plays a role. As Elkins noted, collaborative robots—or cobots—have become all the rage over the past few years, mainly due to their ease of use and seemingly safer operation in proximity to humans. Yet appearances can be deceiving. Cobots can be slower and less accurate than so-called “industrial robots,” possibly slowing ROI, and even the friendliest cobot can harm a human if established safety procedures aren’t followed.
This is why Elkins and others strongly recommend a risk assessment with any robot installation. He also suggested that the best approach for automation newbies is to start small and easy. “Find a single pain point and conquer that one first before looking at adding additional processes or complexity,” he said. “You want to have success right out of the gate so that everyone—the shop owner, the machine operators, and the people doing the integration—are comfortable moving on to larger or more complex projects. Sometimes, the biggest challenge is getting employee buy-in. There’s always the fear that robots will replace humans, so it’s important to make them part of the implementation. Educate them, get them involved in the roll-out, and when they find that the robot actually makes their job easier and more interesting, they’ll end up being your biggest cheerleaders.”