Product development at Canon can be characterized in three stages. First, there is applications-oriented basic research. Middle managers establish potential product applications at a very early stage in basic research. Projects are reviewed frequently to determine whether applications are realistic. If viable product concepts cannot be defined within the first 12-18 months of basic research, a development effort will be refocused on more productive areas. Product concepts are not simply general ideas for applications, but are outline specifications of product features and functionality, with detailed assessments of the impact of a technology on a particular product. At this stage the current needs of existing customers are not considered. In the second stage, a single product target is selected from a number of potential applications of a new technology. Substantial resources are concentrated on that target, and a development team is created to bring that single product to market. Resources are swiftly focused on priority areas, and rapidly shifted between areas when priorities change. This resource shifting is illustrated by the activities of the Component Development Center at Hiratsuka. During the 1980s, CDC's focus was shifted four times on different technological targets. Although small projects were developed in parallel at the same time, the organization focused overwhelmingly on a series of major priorities. Successful development of bubble jet technology was followed by focus on the BASIS device for autofocus cameras, which was followed by development of contact sensor components for fax machines. At the end of the 1980s, and then into the 1990s, the facility shifted its focus once again to develop flat-screen technologies. The third and final stage is diversification. Once a specific product has been successfully developed and has been brought to market, the technology will be applied to other products. As a general model, Canon aims to take a single product to market as quickly as possible, and then to apply the technology (with appropriate modifications) to other products. Over time, Canon develops superior core technologies that can applied to a number of different products. For example, bubble jet technology was initially developed as an application for digital copiers. When this attempt failed, Canon rapidly applied the technology to printers. Once a bubble jet printer was successfully developed and taken to market, the technology was then applied to other products, for example, fax machines and laptop PCs. Below we will explain Canon's development of contact sensor from application-oriented basic research to the development CS-I product and then CS-II product. Phase 1. Application-Oriented Basic Research In 1975, two scientists at the University of Dundee in Scotland demonstrated the special photo-conductive qualities of amorphous-Silicon (a-Si). At Canon, a number of engineers realized the potential applications of amorphous-Silicon technology in imaging products. In 1977, Mr. Toshiyuki Komatsu, a young engineer at the Canon Research Center, joined a team at the Tokyo Institute of Technology, which was conducting basic research into a-Si. Komatsu's task was to study the material and develop a feasibility study of applications using a-Si as a photo-receptor. Within a year, he had identified three potential applications for the technology: a photo-receptive drum for copier machines, a thin-film transistor for liquid crystal displays, and a contact sensor for digital copiers, fax machines and other imaging equipment. Phase 2. Development of CS-I Product In 1978, Canon's a-Si research was organized into three teams which aimed to develop these three applications of the technology. Mr. Komatsu returned from the Tokyo Institute of Technology to lead the team researching the contact sensor. His team had two specific aims. The first was to continue basic research in order to develop a defensible patent position in the technology which they achieved by applying for 128 patents between 1978-1982. The second aim was to develop a detailed product concept for a contact sensor. Komatsu's cross-functional research team (including members from the Canon Research Center, the Component Development Center, and from the Production Technology Research Center) held many offsite brainstorming sessions. These offsites produced a detailed product concept for Canon's first contact sensor, the CS-I, and developed a functional prototype to demonstrate its technical feasibility. By mid-1984 the elemental prototype of the CS-I had been developed. Upon successful completion of trials, mass production of the CS-I started at Hiratsuka in Mid-1985. The final stages of product development were extremely intense, as the joint development team attempted to iron out problems with the component. Production was difficult because a-Si remained a relatively unknown material and its behavior in the production process was unpredictable. The engineers designed the production process around problems encountered with the material as they arose. At the end of 1985 the Canofax 110, the worlds first compact fax machine, was launched into the market. This represented a significant technological achievement for the development team, as Canon was the first company to have successfully developed an a-Si contact sensor. The Canofax 110 was received reasonably in the market. But the CS-I component was plagued by a number of problems. Most importantly, it was more expensive to manufacture than competitor CCD system products. Contact sensor technology was potentially superior, but competing manufacturers had come down the experience curve, giving them higher manufacturing yields and consequently lower costs. Phase 3. CS-II Product Concept It was clear to Mr. Komatsu and the development team that they had to produce a technological breakthrough to provide superior functionality and to reduce the cost of the component. In one week in January 1986, Komatsu's team outlined the technology concept and product specification for a second generation contact sensor, the CS-II. The team had experimented with many potential technology options during the long incubation period of the CS-I product concept between 1978 and 1982. This meant they had a good understanding of the advantages and disadvantages of each option and could make rapid judgments about the viability of the CS-II during this intense period of product design. The concept for CS-II was a significant technological leap ahead of the CS-I. First, it was to be a totally lensless contact sensor. Second, all functionalities would be integrated on the same circuit. The photo sensing elements, thin film transistors, and light guiding elements were all mounted on the same substrate. These improvements would reduce the component size further and could potentially lower the cost significantly. At the end of January, the CS-II Task Force of 20 engineers was formed. The Task Force aimed to assess the feasibility of the CS-II concept as quickly as possible. This period of the project illustrated the value of frequent offsite discussions: difficult technical problems were discussed by the whole team, generating a large number of ideas and solutions. Task Force, Task, and Project A Task Force at Canon is a small team of 5-20 members, brought together for a short-term period, to develop a product concept and conduct a feasibility study. A budget is not formally allocated, but is usually 'bootstrapped', or contributed out of another project's resources. A successful Task Force may grow into either a Task or a Project. The purpose of a Task is to develop and establish a new technology. A Task has a longer time frame, about two to three years, is staffed mostly with full time members, and has an allocated budget. However, a Task does not appear on the R&D organization chart. Approval of the Research and Development Center Heads is required to set up these Tasks. In contrast, the purpose of a Project is usually to start a new business area, or to launch a new product. A Project is a long term, organized structure, staffed completely with full time members, that has an allocated budget, and that appears on an organization chart. To set up a Project requires the approval of the Executive Committee. In April 1986, the team produced an elemental prototype that confirmed the technical possibility of the idea behind the CS-II, but many issues remained. Organizational Problems The CS-II was a significant technical advance that required new manufacturing skills. Canon had never before attempted to produce a complex thin-film component such as this over such a large dimension. Despite their general manufacturing experience, it was not clear how quickly the production team could develop new technical competencies in this area. Development of the CS-II would also have significant implications for project management. On this particular project, development and production engineers were involved in a major improvement program to support existing customers of the CS-I and were struggling to improve 'QCD' Quality, Cost and Delivery of the component already in the marketplace. To add the development of the CS-II could over-extend the management and senior engineers on the project. Cost Problems Manufacturing costs were highly uncertain. The key to component cost was process yield, which was dependent on the width of the component and the cleanness of the production process. Sixty-four contact sensor components were produced on a single plate and were then separated at the end of the process by diamond-cutting. The number of defects per plate, due to contamination from dust and other particles, was fairly constant. This meant that if the width of each contact sensor could be reduced, and more could be fit on a single production plate, a smaller percentage of the total number would be damaged by contamination, and manufacturing yield per plate would improve. Therefore, the aim was both to reduce the width of the component and improve the cleanness of the manufacturing process. However, maintaining the functional reliability of the component as its circuit structures became finer was a significant challenge. There was much discussion about how far costs might eventually be reduced. Through an analysis of published technical reports, the team could estimate the current cost, and potential future reductions in cost, of competing CCD systems. The cost of the CS-I was known. The team then estimated the potential cost of CS-II. This was highly speculative as production yields could only be approximated. However, the team concluded that although CS-II costs would initially be higher than those of CCD systems, they could potentially be reduced at a faster rate, and would eventually become lower than CCD costs. Market Problems There was also significant consumer and market risk in the decision. The fax market in Japan was very large because the use of faxes for business developed earlier and faster than in most other countries. However, this market was dominated by CCD system fax machines. A CCD system fax is box-like in shape and is bulky and heavy. By comparison, the contact sensor component fax machine is extremely compact. However, at this time consumers were content with existing CCD technology. Fax machines in 1986 were regarded as office machines, not as personal technology, and their large size and lack of portability was not an important functional problem for consumers. The compactness of contact sensor machines was not a characteristic that consumers valued. The team produced a speculative estimate of the potential market for fax machines using contact sensors. However, the fax marketing and business development group, responsible for selling Canon fax machines, were excited about the potential of CS-II technology. Canon was relatively new to the fax market, and the CS-II gave the company a potential basis for product advantage. At this point, the management team was divided over the decision to invest further in the development of the CS-II. Dr. Mitarai was faced with a large number of uncertainties as he thought about his decision. Lets help Dr. Mitarai by answering the following questions.

2. Canon distinguished between three types of need for teams: Task Force, Task, and Project. In class, we discussed four types of structures for new product development teams. (1) Which type of team structure is appropriate for Task Force? (2) Which type of team structure is appropriate for Task? (3) Which type of team structure is appropriate for Project? (4) One organizational problem for Canon in developing CS-II is that it could over-extend the management and senior engineers on the project. From the perspective of team structure, can you provide one solution to this problem?