12.12.05
Implant Manufacturers, Robotics Are Finding Common Ground
Automation could be the answer to higher material costs, lower reimbursement
Frank Celia - Contributing Writer
In the past, robotics and automation have not been major factors in the orthopedic device industry or most other medical device fields. Volume was considered too low, and the design parameters were so complex and exacting that it was thought only the human touch could produce first-rate, quality medical devices.
However, attitudes have shifted over the past five years among more manufacturers who are turning to automation to increase efficiency, enhance safety and reduce per-unit production costs.
In the orthopedic industry, this change is being driven by the increasing demand for products—especially knee and hip implants—which has caused manufactures to ramp up output. At the same time, reimbursements continue to drop while the cost of raw materials is on the rise. Increased product output coupled with decreasing profit margins means device makers are looking to cut waste and overhead costs on the factory floor—and robots are a terrific way to accomplish this goal.
For their part, robot cell manufacturers and integrators are casting a favorable eye toward the medical device, biomedicine and drug discovery industries, mainly because these businesses are showing consistent growth and passed through the most recent economic downturn largely unscathed. “Robots in medical devices have experienced aggressive growth during the past five years,” told a marketing manager for robot manufacturer Staubli Corp. to the e-magazine Robotics Online. “Robot manufacturers are looking for places to sell, and the drug device and pharmaceutical industries haven’t seen much of a recession.”
Orthopedics is performing especially well. The industry has experienced double-digit growth in the past several years, and experts predict it will continue to do so, largely because of the aging Baby Boomer population. Also, the quality of orthopedic implants has improved to the point where what was once seen as a medical option of last resort is now considered a successful and beneficial type of surgery. It is becoming more common for patients in their 40s to receive orthopedic implants and still enjoy beneficial clinical outcomes. Some trade experts expect the orthopedics industry to triple in size by the year 2025.
Given such growth potential, the orthopedic device world is likely to deepen its acquaintance with the robotics world, and vice versa. Here is a look at the role automation plays in manufacturing and how it may develop in the future.
Buffing and Polishing
A robotic cell polishes and buffs orthopedic implants. Photo courtesy of FANUC Robotics and Acme Manufacturing Company. |
Prior to 1990, almost all buffing and polishing was accomplished manually. The process is extremely labor intensive, and the work requires great skill. It takes six months to a year to train the average employee, and the job is tiring, dirty and dangerous. The risk involved means there is a greater chance of incurring workers’ compensation claims and lawsuits. The worker must constantly engage in a repetitive motion, increasing the chances of developing carpal tunnel syndrome.
To do the job, the employee stands before a buffing wheel, which is about a foot in diameter. The buffing lathe spins extraordinarily fast, usually between 1,600 and 2,300 rpm. The speeds are so fast that if the employee is not careful, the part can be ripped from his or her hands, causing serious injury.
Since the late 1990s, most, if not all, of the large implant manufacturers have switched their buffing facilities to some form of automation, according to Fritz Carlson, president of Acme Manufacturing, Auburn Hills, MI, an integrator of robotic buffing technologies. But a few smaller manufacturers still perform this task the old fashioned way. Carlson estimated that in any given year, orthopedic manufacturers represent between 15% and 30% of his company’s business.
Carlson was present at the vanguard of this technology’s entrance into the orthopedic world. In 1990, representatives from Zimmer asked his company to help them develop a polishing and buffing automation system for their artificial knees. At the time, Zimmer had purchased a robot cell from the Japanese company, Yamaha, which had designed a robot system to polish and buff its commercial musical instruments. Zimmer employees had hoped to train the robot to perform buffing and polishing tasks on its medical devices but experienced problems working out the bugs. “We worked with them, developed the tooling, the finishing head technology, the software, the programs—all to generate that mirror finish,” Carlson said.
The importance of a smooth finish is paramount, especially in knee implants. “It has to have a bright finish; it has to look like a piece of jewelry or even better,” Carlson noted.
Bob Penque, president of Pinnacle Technologies, another robot buffing integrator, based in Wyckoff, NJ, agreed.
“A knee is functional, not decorative, like a plumbing fixture,” he said. “When it rides on a high-density polyethylene surface of a tibula tray, if that surface is too rough and not polished correctly, it can wear out the polyethylene and cause premature failure of the knee joint.”
Penque, who says about 20% of his clients are orthopedic manufacturers, noted that these robots are a vast improvement over manual labor.
“The dimensions are generally better than those that can be achieved manually,” he said. “Also, eliminating manual polishing reduces the cost of the number of people involved and improves product quality.” In addition to saving on labor costs, manufacturers also save money on insurance and paying for workers’ compensation due to injuries.
Consistency is also an important benefit. “What you train the robot to do, it does the same way every single time,” Penque said. “You can’t get that from a human.”
The rule of thumb is one robot system can do the work of 10 to 15 people. “So the return on investment is tremendous. In less than 12 months, the system should pay for itself,” he said.
Finally, modern robotic cells are flexible enough to allow for processing of different types of orthopedic components, according to Carlson.
In other words, a cell dedicated to polishing knee parts can be reprogrammed to process hip parts.
However, the robot cells are not fully autonomous. It takes one operator to supervise every one or two cells. “Robots, as precise as they can be, don’t have eyes,” explained Penque. “You need someone to double-check what the machine is doing.”
Off-line Programming
Off-line programming has been popular for robotic applications in venues such as the automotive and aerospace industries for more than a decade, but it is still a recent development in buffing and polishing automation, mainly because of technical difficulties posed in writing the programs, explained Penque. “You have a flexible contact wheel, for example, when doing robotic abrasive belt polishing,” he said. “So it really did not lend itself easily to being programmed off-line, because you have to take so many different flexible parts into account.”
Put succinctly, off-line programming simply means the robot cell can be taught to perform a task via a computer while it continues to perform a separate task. Without off-line programming, workers must stop production, enter into the cell and manually walk the robot through the motions required to perform a task. Using an electronic keypad known as a “teach pendant,” operators “teach” the robot what to do.
“If you are teaching your robot, that means you are not running production with that cell,” said Carlson. “On the other hand, if you can bring in this teaching information electronically with your computer, you can essentially begin to generate parts off-line.”
Going into further detail, Carlson discussed the advantages of off-line programming. “From our point of view, it is a tremendous sales and engineering tool. If we have never set up a particular kind of system before, we can actually produce a three-dimensional simulation of what we think the system is going to look like. It doesn’t matter what language you speak; if you watch the simulation, you can see exactly what the process is going to entail,” he said. Carlson added that the software enables the process to be pre-engineered at the proposal stage.
According to Carlson, almost all of his company’s clients have purchased the licensing agreements to at least one off-line program software package. “Years ago, manufacturers would have to dedicate an entire cell just to generate teaching programs,” he said. “Now they can buy an off-line programming package for a fraction of what it would cost to buy a teaching cell.”
Foundry Applications
The outside of the robotic cell that’s designed to polish and buff orthopedic implants. Photo courtesy of FANUC Robotics and Acme Manufacturing Company. |
The parts involved in orthopedic implant devices have such complex shapes and finishes that it is practically impossible to produce them by traditional machinery techniques. Hence, manufacturers must first mold the part out of a stable, wax-like material. The wax model is then dipped in a ceramic-like slurry until it is completely coated. When the ceramic-like material dries, it hardens into a solid shell surrounding the wax molding. In the next step, the wax-like interior is heated until it liquefies and is drained from the hardened shell. Molten metal is then poured into the area vacated by the wax-like material. When the molten metal dries, the outer shell is broken and removed, and the process has yielded a metal part.
Traditionally, these steps were performed by manual labor. But with robot integration offered by U.K.-based VA Technology, this process can be automated. The primary advantage of automation is that greater consistency can be achieved, said VA Technology president and founder Jim Byrne. “The man in this process is literally holding the part and dipping it into rotating vats of ceramic slurry,” he explained. “No two workers are going to dip the part in the exact same orientation each time.” Byrne added that each step in the process offers room for variation, whereas automation allows the exact same part to be produced every time.
In addition, automated foundry applications are exponentially faster than human workers. It would not be outside the realm of possibility for a robot cell to perform 10 times faster than a team of humans, Byrne estimated.
“In the orthopedic industry, there is clearly a demand for high levels of quality, traceability and consistency,” Byrne concluded. “All of these demands are typically better met by machines than by humans.”
A Vision of the Future
Most integrators lament that robots are unable to “see” as a person does. However, they know technology-endowing robots with a limited type of “vision” are possible to achieve. Carlson’s company, for example, is working on a system that could give robots the ability to inspect parts. “That is a direction we are moving in,” he said. “If [more advanced vision systems] could be developed and used in a production environment, they would offer many advantages. We could fine tune the process even more.”
In fact, other industries outside the medical manufacturing field have benefited from specialized machine vision systems that can aid a robot in “seeing” an object before or after it handles it.
How this works varies by application, but essentially a vision camera or sensor is located in a fixed position above the robot’s workspace, or directly to the robot’s end-of-arm tooling. This gives the vision sensor a good view of the specific part. The vision device is calibrated to the robot’s working coordinate system. From there, the vision system is trained and used to locate the object in the robot’s workspace and report the location (either in two- or three-dimensional space) to the robot controller. The parts can be presented as static or moving, depending on the application and industry.
Many of these technologies have been around for years, even decades. Pre-eminent robot manufacturer FANUC Robotics offered its first vision system back in 1982. However, the technology has found limited popularity in the drug and medical device industries.
Apparently, some industries are reluctant to embrace automated vision systems. “One of the reasons is that the technology is still plagued by its past failures,” explained Edward Roney, FANUC’s manager of vision product development. “It is no secret that machine vision was oversold in the early 1980s. Practical experience did not exist, and the technology was exciting and promising, but young. Many first adopters became disappointed and are still hesitant to the trust the technology.”
A little hesitation is not always a bad thing, Roney added, because high expectations of what the technology can accomplish should be tempered by realism, “especially with the large variety of vision systems available today—from low-end sensors to premium 3-D laser based systems.”
Roney believes attitudes toward this technology are changing. “Three elements today encourage greater adoption of vision-guided robots,” he said. “Improved awareness, lower costs and higher reliability.”
Vision systems are a quarter of the cost they were just 10 years ago, Roney pointed out, and the engineering cost of integrating the systems has also decreased significantly.
Tools such as geometric pattern matching, in which the features of an object are identified and matched instead of identified on a pixel-based comparison, have significantly added to the reliability of machine vision, he stated. “For industrial robots that work in manufacturing plants making parts or products from raw materials or in-process goods, this reliability has greatly increased the applicability of machine vision and reduced the maintenance concerns once associated with previous systems that were tremendously light-change sensitive,” Roney said.
Like other industries before them, orthopedic device makers are in the process of meeting the challenges and reaping the benefits of adopting robotics and automation into the manufacturing process. As demand and overhead continue to rise, this should bode well for profit margins, future growth potential and the patients whose health depends on the finished products.