A multifunctional team with a set of clear goals helped IBM design and manufacture the versatile and low cost Proprinter
In 1981, IBM had sold PCs at a rate far beyond its initial forecast of 250,000. At that time, IBM's least expensive printers sold for $5,000 and were designed for use in data processing centres. The low cost printer market was completely controlled by offshore manufacturers centred in Japan. IBM had two options: to continue...
Low-end printers had too many parts and too much labour content to be profitably built in the United States. James T. Vanderslice, the vice president of IBM for manufacturing and development, was not happy with this conclusion. He told his team: "Let's suppose you didn't work for IBM. Could we do it then? If we can't do it, we're out of the business." This challenge persuaded IBM to find a way to become a competitive manufacturer of low-cost printers in the US. A five-member team was commissioned by IBM to establish design concepts for a low-cost printer. The team consisted of both engineers and marketers to establish basic product objectives. They disassembled competitor's products to get ideas and analyse the competition. They were surprised to find that existing low-cost printers consisted of 150 to 200 parts. Furthermore, only about 50 of these were working parts. The rest were rivets, screws, fasteners, cables, belts, pulleys and springs. These parts were impossible to assemble robotically, so they had to be put together manually. Concurrent engineering utilises multifunctional teams and parallel development to reduce the time to market a product. Rather than developing the product linearly (one step at a time), various tasks were performed simultaneously. Marketers, designers and manufacturers worked together to form multifunctional teams. Together, they were able to anticipate problems and bottlenecks and work them out before delays occurred in the project. In order to compete with foreign manufacturers who used low cost labour, IBM decided to automate the assembly process using Design for Automated Assembly (DFAA). These rules are similar to Design for Assembly (DFA) and Design for Manufacturing (DFM) guidelines. DFA forced the design team to follow guidelines making robotic assembly easier and less expensive. Charley Rogers stated another advantage of DFAA: "Automation drove the integrity of the design. I believe that robots were essential to instil the discipline to do it right; because the parts could not be touched by human hands, it forced us to do it right." The size and weight range of parts were monitored to allow for assembly by a single robot. Each part was firmly attached by a robot leaving no loose ends to be held down for later assembly operations. The Proprinter design team removed all extension springs from their design. They found that they still needed to apply pressure on the paper roller. To fill this requirement, they developed a cantilever spring that was moulded into the side frame. This eliminated all the handling and assembly problems while maintaining functionality. The cantilever springs that hold the paper roller also have bearings mounted onto the end to allow the roller to spin freely. Two major strategies helped hold down the number of parts. Whenever possible, parts with different functions were combined together so that a single part performed multiple functions. A good example of multifunctional parts are the side frames which replaced 14 parts. The second step was to use snap fit fasteners which were moulded into other parts. This eliminated nuts, screws and washers, further reducing the number of parts. In order to reduce costs, plastic parts were incorporated into the Proprinter design. Polymers are generally significantly less expensive than metals. Plastic can also be injection-moulded into complex shapes allowing the design of multifunctional parts. Six types of thermoplastic moulded material were used in the design. Each different type of plastic used had different material characteristics. The snap fits were made possible because of the chosen plastic's flexibility while the side frames were made of more rigid plastic containing carbon fibres for stability and static discharge. Plastic parts could be injection moulded at the plant. This helped reduce the While In Process (WIP) inventory of parts because plastic parts could be made as they were used in the assembly process. Ultimately, 77 per cent of the Proprinter parts were made of plastic. Complete automation of the Proprinter assembly was achieved. The robots were purchased from another branch of IBM who acted as a vendor to the Proprinter group. Each robot performed multiple assembly functions as the printers moved down the assembly line. Only one manual adjustment in the entire assembly process was required; otherwise no human hands were needed on the production line. The automated assembly success was a direct result of good design practices. Standard testing cycles at IBM consisted of four sequential test cycles that took years to complete. Furthermore, certain lengthy tests had to be successfully executed before any manufacturing could begin. The Proprinter team worked with an independent team from product assurance who concentrated on quality. Together, they compressed the testing cycles so that they overlapped and didn't hold up manufacturing. Vendors were required to send process variables via modem to the assembly plant before shipments of parts could be made.
buying low-end printers from overseas or to design and manufacture their own. After a manufacturing analysis, it was evident that IBM could not build a competitive low cost printer.
Compiled by Ajith Ram
(c) 1998 IBM
Read more on Data centre hardware