by [TC]²
A monthly column of technology rambling, rumination and reality
By: Jud Early, Corporate Vice President, [TC]²
April 2009
Deeply Rooted In Technology
Part 1
It is said to be a sign of old age when one reminisces about the good old days. My reflections on the past are not so much a desire to return to those days, but an expression of amazement at how far we have come during the life of [TC]². In order to reduce an oral history to something more permanent, I am writing today about the technology roots of this company and the path along which we have advanced.
Many readers of this article will have little memory of 1980. The big three auto makers were beginning to adopt robots in the production of automobiles. Personal computers did not exist, except as hobbyist devices, usually a bare printed circuit board with an audio tape recorder for program storage. Land lines for communication were the norm, with satellite communication only within the realm of broadcast programming or high cost satellite phones. Cell phones were not yet a consumer item. Radiotelephones were “car phones” of the day.
The men’s tailored clothing industry was largely domestic, with a few Italian suit labels offered in high end department or men’s stores. The threat of imported clothing was just beginning to be understood when the original founders of [TC]² met to organize a combined research effort to counteract the potential harm if the imports were allowed to remain unchecked. A study funded by the National Science Foundation, conducted by researchers from Harvard had estimated the damage, and had offered a technology solution to counteract the threat of imported clothing. Money was pooled and a committee formed to guide the research. I’ll spare the details of how we were organized for another day.
If Detroit could use industrial robots to paint and assemble automobiles, why could not the same robots be utilized to replace costly labor in clothing manufacture? A consortium was formed, with textile manufacturers, clothing manufacturers and organized labor support for mounting a project to research the use of robots. Not to be a backyard operation, the Draper Laboratory in Cambridge Massachusetts was engaged to perform the research, and to develop the first robotic sewing machines. Draper was a lab that was engaged in military contracts and defense projects and was not lacking for highly educated scientists and engineers. What was lacking was knowledge of sewing, and of garment construction in general.
These skills were learned over a period of a couple of years, while obstacles to success were being identified. Vision systems were rudimentary, so the use of security cameras and custom developed software were required to allow the machines to “see” what to pick up and sew. Fabric pick up was another area in which technology had not been adequately developed. Cluett Peabody had developed the Clupicker for shirt fabrics, but its use on a robotic arm was not practical. Another obstacle, another solution. The Walton picker was developed, a simple device that was both inexpensive and compact, a feature that allowed close spacing for multiple pickers that would be required to pick, place and fold the parts to be sewn. End effectors, the robot equivalent of fingers and hands, were developed to allow the pickers to grasp the fabric by its face and by manipulating the end effector over a vacuum table, folding of fabric parts would be done, with further placement and release to a specially designed multiple belt mechanism that carried the parts to be sewn through a specially modified sewing machine.
The resulting machine was demonstrated to be capable of picking up discrete fabric components from a shingled stack, folding, aligning and sewing into a seamed jacket sleeve. The same technology was later to become an assembly machine for men’s trousers. Stretching to over fifty feet in length, the machines were very heavy, costly and with the robots of the day, inflexible due to limited programming controls. Simultaneous to the development of the jacket machine, was a gantry robot for the production of jeans front pockets. A multi-station machine, the right front pocket facing was presented with a watch pocket for the right side, and after folding to conceal raw edges, the two were sewn together. The facing, along with its watch pocket was joined to the pocket bag before closing the bag. This machine was a much smaller machine, occupying an area of about ten feet by fourteen feet. Again, the Achilles heel was programmability and the need to change frequently when different styles were in manufacture.
Other developments built off of the basic understanding of fabric handling and robotic sewing. An attempt was made to develop a fully automatic inseam machine for denim jeans. With two sewing heads and followers to tension the seam, two leg panels were loaded, and were guided by sensors to ensure filling of the folders that created a felled seam. The machine worked, but loading time was too great, and the productivity was too low to justify the cost of manufacture. There was benefit from this early prototype. It was known that stitches are not always properly formed by the sewing machine, so the need for a skipped stitch detector was identified. The success of the sensor guided folder to ensure complete filling of the folder and assure quality led to the development of a semi-automatic felled seam machine, developed under license by AMF Reece, and sold in some quantity to jeans manufacturers who were still in the U.S. The skipped stitch problem was resolved by another sensor that was able to discern the length of thread used in a single stitch, and when a stitch was skipped, resulting in less thread consumption, a signal was sounded and the operator alerted. This was licensed to Quick-Rotan for manufacture and distribution.
A basic requirement that stood until the mid-1990s was that technology developed by [TC]² or its contractors could only be licensed or sold to companies in the United States, a tenet that goes back to the original purpose for our founding, the protection and competitiveness improvement of the domestic industry. With this restriction, equipment commercializers could not sell outside the U.S., while at the same time the apparel manufacturers were moving offshore.
Another machine with origins at Draper Lab was the knitwear pants machine. Originally intended to sew the inseam of fleece warm-up pants, commercialization efforts were licensed to JetSew, formerly a research center for Cluett Peabody, and at the time a part of Fruit Of The Loom Corporation. With a good complement of bright people, experienced in sewing of shirts and briefs, the knitwear machine quickly took shape, and was tested at Russell Corporation. Fruit of The Loom also had interest in the machine, and wanted to use it in offshore operations. This desire coupled with the potential of a large number of machines to be placed in service provided the impetus to obtain the OK to allow machines developed by [TC]² to be used outside the U.S. as long as they were being operated by U.S. interests. The machine performed well, and new models were introduced. The original had been designed to pick fabric plies from a loaded supply cartridge and align them on a table before sewing. The machine had to adapt to mis-cut parts, parts that were wrinkled, sometimes folded, and to cutting tolerances that despite die cutting were inconsistent. The new machine was to be fed from a roll of fabric, and would cut two panels in a single ply cutter station that was a part of the machine, and would then align and sew the pants into a pair, requiring only a bottom finish and waist completion. This machine was running well in test, but as the principal customer began to move offshore, the decision was made to not automate offshore factories where labor was expected to be cheap. The machine would never reach its potential.
The last project in major equipment development was done in concert with Gerber Technology. The discovery of a laser system being offered by General Electric that utilized lasers from Fanuc of Japan formed the basis for development of an ultra-high-speed laser cutter that cut fabrics at over 200 inches per second. With 1500 watts of CO2 power, cutting was visually impressive.
Through the use of a scanning, translating parabolic mirror, the fabric cutting machine was faster than any laser cutter known at the time. The market for such a machine was to be automotive air bags, where the edges being sealed by the heat of the laser was a benefit, not a detractor. In order to install the machine in our demonstration center, a hole had to be cut into the roof, and a “penthouse” built to accommodate the height of the machine. With a conveyor to bring fabric into the cutting chamber, and a speed so fast that all that was visible was a “rooster-tail” of flame as the fine beam burned through the fabric, the machine was impressive to watch in operation. Interestingly, during its time in our demonstration center, it was used by an outside developer to laser etch denim in one of the first experiments in ablation to remove the color from fabric without harming the fibers. Patents were granted as a result of the early research. Two machines were built, but just as apparel had done, the production of air bags was on the march out of the country.
All the elements had come together, the laser and mirror technology, the need for sealed edges, and a market study that indicated a market of good size would make the project financially justified. The element that could not be measured was the speed at which the need evaporated, leaving the cutter with no place to go.
You may be asking what the development of all this old equipment has to do with where we are today. [TC]² is nothing if not agile, changing as the industry has changed, and adapting its technology agenda to meet the needs of the industry of today. Technology was key to our founding, and has been a central part of what we do for more than twenty-eight years.
This is the end of part one of this two part series. In the next part we will review how lasers figured into our training products, and how they did not find favor in our body scanner development.