Flexible cells adapt to meet future demand
March 7, 2013
When a robotic manufacturing system needs to be moved, particularly between remote sites, the cost of the move can be a significant percentage of the original cost of the system. But this doesn’t always have to be the case.
Demand variability is an immutable fact of contract metal fabrication. A fabricator’s customer mix can change and, with it, the product mix. Because of this, the shop has to be ready for anything. To be competitive, many have rethought traditional concepts of shop layout to reduce travel time between processes. If the product mix shifts significantly, shop managers may want to rearrange machines to suit.
This isn’t such an arduous affair for small machines—but what about robotic welding cells? Most don’t consider moving them, and for good reason. Moving a complex cell requires serious planning and significant costs, which traditionally has made the whole endeavor impractical.
Even so, to be competitive today, a manufacturer must produce a variety of high-quality products. Short product life cycles coupled with diverse product portfolios creates a high-product-mix, low-volume scenario. And as production demands increase, manual methods are becoming less viable. In short, automation is no longer an option. It has become a necessity.
Two driving forces push companies to automate: increasing throughput and maintaining consistent, high product quality. To maximize throughput, robots traditionally have performed repetitive tasks that don’t change over time and have short cycle times.
But the question remains: How does a company embrace automation and remain adaptable? Companies must find ways to make systems as flexible as possible. For instance, many now use tooling and fixturing that can be changed out between production runs of different products. In this so-called “picture-frame” assembly, a permanently mounted frame inside the positioner allows different fixtures to be mounted and unmounted when required.
Still, true challenges can arise when they need to move entire robotic cells from one area of a facility to another, or even to another facility altogether, to keep pace with the demands of an ever-changing production environment.
Large robotic cells and production lines involving multiple mechanical units generally are not easy to relocate to other areas in a facility or between facilities that may be hundreds of miles apart. Such a move involves many costs: labor, utilities, freight, reinstallation, reprogramming or recommissioning, and operator training—to name just a few. When a robotic manufacturing system needs to be moved, particularly between remote sites, the cost of the move can be a significant percentage of the original cost of the system. Often a more economical option is to apply those relocation costs to a new system.
This doesn’t have to be the case, however, especially if a welding robot cell is initially designed to be relocated, should the need arise. Every robot system requires certain connections for fume extraction, system power, and other elements, so relocating any cell requires planning. But some welding robot cells are easier to move and reconfigure than others. Flexible robotic systems come in a variety of configurations, but in a broad sense, they can be grouped into two categories: modular and palletized.
In a modular robotic system, not every component is permanently hard-wired to other components. Instead, certain elements in a modular cell can be disconnected and reassembled into an entirely different configuration—one with a bigger or smaller work envelope, a different positioning or fixturing system, and so on.
These advanced features, however, require significant forethought. During the earliest design and planning stages, everyone must have a clear understanding of the product mix, production system flow, and how best to achieve the desired return on investment. Fabricators that must deliver optimal part flow now, yet retain the ability to adapt to future product-mix changes, should consider the advantages of modular work areas from the start of the planning and design phase. It isn’t something that can be adjusted easily at the time of installation.
Such a robot cell is extremely adaptable. Instead of having a work area dedicated to one task, the cell can have modular units that can be swapped out to accommodate any type of device, from simple tables and clamping units to headstocks and small, multiaxis positioning units.
These robotic cells can be designed to handle significant product variation. Advanced features can simplify relocating a robotic cell from one part of the plant to another. But all this can add engineering costs during the design phase. Therefore, companies should conduct due diligence to determine if they can attain the required return. High-product-mix, low-volume operations typically have a lower rate of return because of the additional time required for start-up, part programming, and lower overall part count.
In fact, for some companies with stringent capital expenditure requirements, the additional engineering and production benefits can be challenging to define purely in financial terms. In this case, it might be better—and less costly—to install additional dedicated cells in lieu of a modular system.
A more economical mobile option is a palletized robot cell. A contract fabricator can install a common base plate (or picture frame) that holds different fixtures. If fixtures are staged, and programs written and tested offline, a technician potentially can switch the fixture, call up the program, locate the part, and start the next job within minutes.
Unlike modular systems, palletized cells usually are designed as a unit with a fixed work envelope. This limits the size and type of parts the robot cell can accept. And even if the part does fit inside the work envelope, the robot arm may not be able to reach every weld joint for every part.
Still, these systems can be moved to another area of the plant—or to a different plant—with relative ease. Large palletized units can be broken down into smaller components and shipped in sections. Some of the palletized cells are even designed with built-in receptacles to enable movement via a forklift. When designed efficiently, they can simply be unplugged, lifted, and dropped on a truck for relocation.
If cell mobility is the goal, either a modular or palletized alternative helps address the need of high-mix production. Know, however, that there are trade-offs between lower-cost palletized systems and more expensive modular systems.
When a robotic cell needs to be moved, shops can take steps to streamline the process. Moves should be planned in advance, never made on-the-fly.
For example, whether using modular cell components or a palletized system, shops must ensure upfront that the units are built with quick disconnects for electric and air cables to enable rapid removal from the utility grid. Gas bottles mounted directly on the units avoid the need to locate the cell near a central gas connection. Local welding wire delivery eliminates the need for connection to a remote wire source. Local fume-extraction units also may be used if no central system is available in the shop. Portable air circulators will accommodate multiple cells.
Safety also is a big consideration in a move. Many modern robotic cells have integral safety systems that are compliant with Robotic Industries Association standards. No additional safety items are required when these units are moved around a facility or between locations.
Robots do not possess human-like intelligence with the ability to reason. A person must program every robot movement. When a cell is moved, mechanical components can flex and move, too, especially when tooling is swapped out.
Because of the inherent slack in mechanical systems, a program that runs flawlessly in one location may have significant errors in another location. Fortunately, several robotic technologies overcome these problems.
These can be simple programming constructs or sophisticated sensors that allow the robot to determine and react to changes in the local physical environment. For example, a technician can perform a 3-D shift on a program to account for fixture displacement, or even define special program frames attached to a fixture. This requires a skilled robot programmer to access the robot program directly and perform these operations.
Instead of investing in expensive and complicated tooling for part locating, shops can implement other technologies that enable a robot to perform these tasks autonomously. For example, touch sensing allows the robot to use either the tip of the wire or a laser sensor to detect basic features on a part and make necessary adjustments to its program. While it is relatively easy to program a robot for this, it can add to overall cycle time.
A faster alternative that requires more programming time during commissioning is to use a camera on the end of the robot. A robot literally can “see” the world around it to perform tasks such as part detection, part recognition, feature detection, and program offsets. It can even detect which type of fixture or tool is loaded in its work area.
The most effective tool at a system designer’s disposal is CAD. By making extensive use of 3-D modeling in the system design phase, the designer can see what a system will look and perform like when it is built. Once a system is built, programs can be written and verified offline, ensuring a system has maximum uptime and effectiveness.
A high-product-mix shop can choose from a variety of robotic welding cell configurations, but generally, these can be boiled down to three options. First, the fabricator can choose a self-contained palletized system. This can be moved easily and is less expensive than other options. But its work envelope is usually fixed, which limits the part sizes and shapes it can weld. Within these constraints, however, these systems can weld numerous short-run jobs that customers repeatedly order.
Second, a fabricator can choose an engineered cell designed to weld a single part or part family extremely efficiently. This can be a good option if the operation has a consistent, relatively high-volume order in the product mix. But these systems cannot be moved or reconfigured easily.
Third, a fabricator may choose a highly engineered modular system, which can be reconfigured for optimal efficiency for various parts. This system can be configured to weld a single part or part family and then be moved and completely reconfigured for an entirely different application.
This last option involves flexibility and redundancy that can require significant upfront costs. For this reason, fabricators need to ask: Do we want to invest upfront in a flexible system that will grow and adapt with our shop, or do we want to amortize such costs by selecting a less expensive, fixed robotic system?
This decision must be made during the design phase because, while flexibility is expensive on the front end, retrofitting a cell down the road can be even more costly. It requires a holistic approach that examines the entire production process to determine the ultimate return on investment.
Shops should look at their processes and procedures currently in place, and then determine what processes they want the automated system to improve. The proposed option for automation must not only make economic sense, but must also fit in with current and anticipated production requirements. As the field of robotics continues to evolve, so will the tasks robots are assigned to.