This section was edited by Associate Editor Alan S. Brown.

ROBOT EXCELS AT LOW-VOLUME ASSEMBLY

Everybody knows that industrial robots boost productivity, but only if you do the same sequence of high-volume tasks over and over again. Now, SmartTCP Inc. of Farmington Hills, Mich., has introduced a robot designed specifically for low-volume production.

"When this thing is on its game, it's absolutely phenomenal," said Tim Bowen. Bowen is production manager at NPK Construction Equipment Inc., in Wickliffe, Ohio. The company makes a family of mounting brackets for hydraulic hammers used in demolition and excavation.

Some of those brackets have as many as 150 weld seams. It took anywhere from 20 to 27 hours to weld one of NPK's largest brackets manually, including mounting, dismounting, and repositioning. SmartTCP does the job in 7.5 hours and leaves Bowen impressed with the quality of the welds.

A new robot developed for low-volume applications
programs itself within hours using a CAD drawing,
then welds the final part.

The problem with robots for short-run production has always been setup. Engineers teach robots how to weld by moving a pendant over a part to show the robot where to position itself, then check, double-check, and reprogram to improve precision. When Bowen was shopping for a robot, integrators told him that it might take two months to program a robot to make a complex mounting bracket. "There was just no way," Bowen said. "The robot would be sitting idle, not producing a single part, while we were teaching it."

SmartTCP does its programming offline, so while the robot is making one part, the welder is programming the next one on the computer. This is possible because SmartTCP can analyze any weldable geometry in a CAD drawing, analyze it, and create its own welding path, according to SmartTCP’s CEO, Efi Lebel.

"If it's a really complex piece, the software will get it 90, 95 percent right. Sometimes there's a narrow space where the robot cannot fit physically, or the system will specify a half-inch weld and you want a five-eighths-inch weld. But we do all the programming offline and you can make any changes in the software. It's very visual, and it can be done by a welder, not a programmer," Lebel said.

Bowen agrees. "The robot gets it right about 90 percent of the time, and takes only a few hours to define the rest of the program," he noted.

Lebel has a very different philosophy about programming robots: "It's like when you take a young guy and teach him to drive. You don't teach him to ride from point A to point B. You teach him to accelerate, brake, and turn. That's exactly what we do. Instead of teaching the robot to weld a specific part, we educate it on how to weld in a specific environment, so it knows what to do when it sees a certain geometry and material in a CAD drawing."

Lebel uses off-the-shelf hardware to build his robots. Although he prefers Kuka robots, he works with other automatons as well. The company's base unit consists of a gantry-mounted 9-axis robot, and a typical installation will use a 12 or 13-axis robot. The company has built gantries as long as 100 feet for fabricating structural steel.

"Because we use off-the-shelf equipment, we don't have to build the system in our facility and then ship it," Lebel said. "The first time we build it, we build it at the customer's site. We can build a system in three weeks and have it working in five weeks. We also offer a one-year guarantee on programming, so if there is a problem getting the robot to build something, we will go over the problem with you, find out what went wrong, and then show you how to do it."

Prices for SmartTCP start at $500,000, but the average system runs about $700,000. According to Lebel, the company has sold eight systems so far. It is focusing on heavy industry and companies that fabricate steel beams, girders, and joints for the construction industry.

PLASTIC AND OIL

A plastics trade show in Germany this fall got a look at a new car part, a 6-liter oil pan module made from a thermoplastic polymer. In what they believe, at least, is a first, three companies got together to work out how to do it.

Daimler AG, DuPont Automotive, and one of Daimler’s suppliers, Bruss, developed a modular design to be included in new four-cylinder diesel engines in Mercedes-Benz C Class cars. The part consists of a die-cast aluminum upper shell and a multifunctional lower shell, made of DuPont’s Zytel 70G35 HSLR thermoplastic.
The design was featured at Fakuma, an international plastics processing trade fair in Friedrichshafen, Germany, in October.

The design has 1.1 kilogram less mass than an entirely aluminum design. According to DuPont, the material makes for higher production efficiency than aluminum. The high flow of heat-stabilized, glass-fiber reinforced nylon enables long flow distances, short injection times, and the reliable molding of thin-walled sections, the company said.

According to DuPont, the rear section of the oil pan, which forms the sump for approximately 6 liters of oil, is very rigid due to its shape. The front section had to be made flatter to fit the chassis and make room for the steering gear, and so is less resistant to bending.

A Daimler oil pan shed 1.1 kg by using thermoplastic instead of aluminum.

The solution was to reinforce it with a sandwich design in which a second injection-molded part, an oil deflector, is welded onto the flat section of the pan. The deflector helps calm the oil churned by the crankshaft and balance shaft, and directs it back into the oil pan.

Brass inserts in the oil sump section accommodate the oil discharge screw and oil level switch. The high ribs in the sump act as baffles, helping to calm the oil and to direct it toward the sump.

Finite-element modeling was used to refine the positioning of ribbing at the edge of the pan to increase the stiffness of the flat section. DuPont said its engineers also did digital flow studies which established that, because of the high melt flow of the material, one central gate was enough to fill the mold cavity completely while permitting short molding cycles.

Physical tests by Bruss confirmed the adopted design. After 1,000 hours of aging in hot oil at 150°C, the pan was able to withstand severe test conditions without incurring critical damage.

ROBOT SEES, ROBOT PICKS

Sorting and picking jumbled parts in a bin is a task that would stump almost any robot, but not the one undergoing beta testing by ABB Robot Automation. Built with an ABB IRC 5 robot controller and vision guidance software from Braintech Inc. of McLean, Va., the system is undergoing pilot tests now. ABB plans to commercialize the technology early this year.

Picking 3D parts from a bin is simple work for even a young child, who quickly recognizes each part regardless of its orientation, determines whether one part is blocking another, decides on the easiest part to remove, and then picks an angle to reach out and remove it. Children do it every day when they sort through toys, clothing, or crayons.

ABB uses a robot-mounted camera to locate 3-D parts in a bin, then calculates the best trajectory to get to them.

As simple as it sounds, it takes a lot of brainpower to make those judgments. Industrial robots have lacked this ability until now. After 30 years, the best they can do is use a single approach path to pick 2-D parts from rough stacks or flat layers in a bin.

Slick engineering and faster microprocessors are overcoming these problems. Today, vision systems can recognize parts and patterns, and calculate approach trajectories fast enough to make random picking possible.

Better vision systems also make a difference, according to Braintech’s chief sales officer, Jim Dara. In this case, Braintech and ABB have mounted two cameras on the robot.

"The cameras are set 6 inches apart. They can see enough of the part to calculate its position in three-dimensional space and determine if it is pickable or blocked by another part. Because there are two cameras, they can triangulate the relative position of the robot and the part," explained Dara.

The robot has several tricks up its proverbial sleeve to manage this in real time. First, it does not start by matching 3-D images. It uses faster 2-D algorithms to match profiles of parts that it might be able to pull. Only then does it move to more time-consuming 3-D analysis to see if it can reach the part.

The robot also knows enough to move around the bin to avoid reflections, irregularities, or just a poor point of view. It can also control lighting to illuminate shadows or reduce reflections. This gets progressively more difficult as the robot works its way to the bottom of the bin. Sometimes, it will jam, then stop and reset without operator assistance.

The system eliminates the need for manually sorting and racking parts for robot use after machining. The result, said ABB Robot Automation vice president Jerry Osborn, is "faster productivity, improved quality, and reduced capital and maintenance costs, all leading to increased profitability."

THINK OF WORKERS AS ASSETS

Hiring the right workers is the single most important action a company can take to improve plant profitability and worker retention. That was the finding of a study of 45 process industry facilities by McBassi & Co., a Golden, Colo., human asset management firm that works with organizations as diverse as manufacturers, schools, and nonprofit groups.

At a time when technology and capital flow with ease around the globe, assets alone rarely set companies apart, said one of the firm’s founders, Laurie Bassi. To remain competitive, companies need to create and properly manage highly skilled workforces.

McBassi's methodology enables large corporations to assess how they manage their workforce across multiple sites with the same type of metrics they might use to look at asset productivity or energy consumption.

McBassi divides workforce drivers of performance into five broad categories: leadership practices, workforce optimization, employee engagement, access to information, and learning capacity. Each sector is subdivided into more specific behaviors. Employee engagement, for example, is defined by job design (work organized to tap employee skills); commitment (employees have secure jobs, and receive recognition and advancement); time (employees have enough time to do jobs well); contribution (organization has ties to the community); and systems (companies continually evaluate engagement).

McBassi then surveys the workforce to see how the company rates in these categories. The result is a systematic snapshot of competencies across multiple plants and divisions.

Some of the findings are surprising. The process industry survey, for example, found that plants with above-average hiring decision scores were 2.5 times more likely to report higher profitability than plants with below-average scores. Plants with high scores on information sharing were seven times more likely to report world-class asset utilization rates.

Of course, those results vary between industries, among competitors, and even within individual companies. This is because companies and practices are evolving all the time. "The key is which subset of factors matters to the organization at this time," Bassi said. "Benchmarks and best practices can tell you what to do, but they cannot help you identify what's important for your company."

MORE THAN NUTS AND BOLTS

Fasteners are still mostly about nuts and bolts. Yet, some catalogs contain products so varied and advanced that it can be hard to recognize them as fasteners, said Dean Peters, an industry analyst with Cleveland-based market researcher Freedonia Group.

"With computer-aided design and analysis tools, engineers can look at how fasteners really fasten, pinpoint stresses, and determine what happens when they vibrate and wear," Peters said. "This enables some of the more advanced manufacturers to create seals that resist vibration better or that you can't over-torque."

According to the study, World Industrial Fasteners, global demand for industrial fasteners will grow 4.8 percent annually to $66 billion in 2012. This appears slower than the 8.8 percent growth rate between 2002 and 2007, but much of that increase was due to runups in raw material prices. Real growth (with raw material inflation factored out) was only 3 percent per year. Freedonia expects real demand to rise 3.6 percent through 2012.

Specialized fasteners remain high-value products amidst the nuts and bolts of the fastener market.

Not surprisingly, growth in Asia, Africa with the Mideast, Eastern Europe, and Latin America will exceed the pace in the United States, Western Europe, and Japan. According to Peters, China likely replaced Japan as the second largest fastener market in the world (after the United States) in 2008.

Through 2012, motor vehicle manufacturers will remain the largest users of fasteners, accounting for about one-third of total demand. "There's some competition to fasteners in the long run from vibration-resistant advanced adhesives, which grip stronger and hold on longer," Peters noted. "We're also seeing more unibody construction, as casting and forging becomes more competitive."

Electrical and electronic equipment, which uses a wide range of plastic as well as metal fasteners, will be the fastest growing market through 2012.