Automation doesn't solve process problems. People do.
October 9, 2007
While automation can increase throughput and reduce labor, it doesn't solve manufacturing problems. A manual process that produces poor-quality or inconsistent parts will simply do so at a faster rate if automated. Understanding the process and process variables is the key to troubleshooting problems and resolving them to get the maximum gain from automation.
Whether it is as simple as a single CNC tube bender loaded by a robot or as complex as a fully automated line that turns raw coil into a finished and packaged bent tubular product, automated workcells have made their way into nearly every manufacturing theater. Once limited to the automotive industry with its ultrahigh production volumes, automated cells now are found in any industry in which the production rate is a "make or break" aspect of cost-effective manufacturing. As raw material costs and labor costs continue to rise and cutthroat competition forces prices down, automated cells are being used to produce more complex workpieces than ever before.
Trials abound. The demand for higher production comes with an inherent requirement that the process runs nearly continuously. Automation in a tube bending cell must be well-planned and well-executed simply to bring the project to a successful launch. Once the workcell is up and running, the maintenance staff must keep it running for two or three shifts a day, six to seven days a week—and that's another challenge entirely.
Successful tube bending is possible only after analyzing the application and matching the application to the proper techniques, machinery, and tooling. If the application is borderline—that is, just barely successful—when accomplished manually, no amount of automation will make it more successful. In other words, automation doesn't reduce problems; it merely increases the output. If a manual process has a 10 percent scrap rate, under the best conditions a similar automated process has a 10 percent scrap rate. Debugging and simplifying the manufacturing process before automating it is critical.
Taken as a stand-alone operation, tube bending is itself a matter of juggling ever-changing variables. To simplify this facet of workcell planning, initial considerations must include:
Also commonly overlooked is raw material consistency. It is absolutely essential that all the tube's properties—the chemical, mechanical, and physical characteristics—be consistent from lot to lot. Make certain from the start that the material selected is toleranced to optimize production, rather than a variable that makes steady, consistent production impossible.
Many tooling design techniques can help meet production demands, but the tooling must be as simple as possible, especially in a high-production-volume environment. Don't throw more technology (and complications) at the application than necessary.
In applications that require several die stacks, make certain that the machine is rigid enough to keep all the die sets in proper alignment. This often requires overhead tie bar supports for the bend die stack. While offering more strength and support, these tie bars also can interfere with some bent tube configurations. Be certain that you and the machine manufacturer understand the requirements of the proposed bent part.
With regard to several stacked wiper dies, remember that these fragile tools must be held in their proper positions securely. Otherwise they will not last or work effectively. It is often necessary to have a custom wiper bracket (holder) made to keep wiper dies secured and aligned properly and thus achieve acceptable die life and greater productivity.
This also prevents manufacturing scrap. Empowering the operators to verify part conformance frequently throughout the manufacturing process enables them to troubleshoot the process when they discover nonconforming parts. Doing so efficiently is a matter of checking parts randomly and frequently while they are being manufactured, not after they are finished.
Because tube bending generally is the nucleus of a workcell, the other processes must be synchronized to this process. Determine the optimal cycle time for bending, and adjust the loader speed accordingly. If the cell involves a shear that cuts the tube or, in an extreme case, a tube mill that produces the tube, these operations must be timed so the bender is neither over- nor underfed.
Be aware that the overall speed of the automated cell or line must be based on the fastest tolerable cycle of the slowest operation of the cell. Also be aware that timing is nothing without debugging. Set up and troubleshoot every operation individually before attempting to integrate them in an automated process.
Sophisticated animation software can facilitate placing and integrating every piece of equipment in the workcell. Most work flow problems and equipment collisions are caught in the programming stage. However, in the real world it can take some time and program tweaking to achieve the mechanized ballet we strive for in a workcell. When the optimal cycle times are dialed in and benchmarks set for speed and other acceptance parameters, the next critical phase of the start-up begins.
Train the Operating and Maintenance Crews. Procedures for operations and maintenance training vary considerably from company to company. Some companies have three separate groups—one for tooling setup, a second for machine maintenance, and a third that operates the cell and monitors production. Considering the cell is likely intended to operate for two to three shifts per day, seven days a week, it is obvious that consistent output among all shifts requires consistent training among all personnel.
Without maintenance training, equipment operators can do nothing more to solve bending problems than find the few knowledgeable troubleshooting personnel and inform them of the problem. Likewise, the tooling setup staff's responses to problems can be limited. "Let's replace the tooling!" is a fast, easy, yet ineffective response if the trouble starts elsewhere.
Ideally, all personnel receive the same training and a set of written procedures so that everyone learns a single approach to troubleshooting.
Keep Accurate, Thorough Records. Equally important—and harder to implement—is an accurate method of monitoring uptime and downtime, cycle speeds, scrap rates, maintenance procedures, tooling settings, and changeovers. This data is necessary for making informed decisions regarding equipment and tooling condition. Adjusting procedures based on this information can help minimize downtime and anticipate catastrophic failure.
Any steps for developing constantly updated (and consistently formatted) records are valuable. The records themselves are invaluable if they are frequently analyzed and used for planning preventive maintenance activities.
If nothing else, good recordkeeping prevents running out of consumable tooling items (wiper dies and mandrels, for example). How many times do you need to run out of $50 wiper tips, causing the multimillion-dollar workcell to shut down and leaving your best customer stranded, before you realize that you should have a supply of all tooling and a steady flow of consumable items on recurring blanket orders? Proper documentation analysis and proactive process implementation are the only ways to prevent consumable shortages.
While it is just human nature to rush to get a high-speed workcell or production line up and running as quickly as possible after a problem arises, it is necessary to trace the cause of it all the way to its source. Follow the process back to the first step that did not meet the acceptance criteria. The principal reason for doing this, of course, is to solve the problem and not merely address the symptom. Allowing the problem to continue unchecked means it will compound later on. Resolving the problem now also prevents repeated and excessive downtime later.
If the workcell is a completely automated line, it might involve a tube mill; straightening, punching, and forming machines; a weld seam detector; a bending machine; hydroforming press; laser cutting system; welding station; and, of course, a material handling system. The complexity and speed of such a system means that a small problem early in the process has the potential to get completely out of hand in the blink of an eye. Constant quality monitoring and proper and pragmatic problem-solving are not just advised. They are required.