Research paper : Development and commercialization of laser inspection system to detect surface aws of machined holes (S. OKADA et al.)−144−Synthesiology - English edition Vol.11 No.3 (2018) inspection technology for glossy and mirror surfaces was developed, because aws and defects were often missed in visual inspections of inner surfaces of holes with various diameters of industrial parts used in automobiles and others. This is a detailed report to show the process and the signicance of using AIST’s technological potential applying semiconductor laser diffraction, in order to develop an innovative defect inspection system, and to succeed in its commercialization. Particularly, this paper mentions that with overcoming crises such as re-organization of AIST and Lehman Shock, it was very effective to conduct product development matching demands of customers. This paper is an excellent paper that thoroughly discusses the scenario for successful product realization.2 Demand for automated inspection technologyQuestion (Ken’ichi Fujii)In Chapter 1, you write, “Visual inspection is done for the inner wall surfaces of machined holes with various diameters that are drilled in cylinder bores, automatic transmission valve bodies, hydraulic cylinders, and others that are important safety parts in automobile industry, since they have particularly strict inspection standards. As no oversight is tolerated, there is demand for high-precision and high-speed automated inspection technology that detects minute aws and defects of about 0.1 mm.” Please provide an easy-to-understand scale that shows the degree of effect of cost reduction through this automation, as well as the amount invested as cost to this type of inspection system by the automobile industry.Answer (Yasufumi Esaki)The cost of the system to realize automated inspection for valve bodies (transmission parts) is 15 million to 20 million yen/system. This system automates valve body hole inspection. In fact, about 60 % of the total items of inspection of valve bodies have become automated with this technology.On the other hand, the normal cost of inspection personnel is about 10 million yen/person/year. Normally, hole inspection of valve bodies is conducted by four people/set, so automation will generate cost reduction of 40 million yen (10 million yen x 4) per year. Of course, improvement of quality was realized at the same time.3 Development of defect inspection system for inner wall surfaces of machined holesQuestion (Ken’ichi Fujii)In Chapter 3, you write that you developed the technology for inspecting the inner wall surfaces of machined holes with excellent ideas such as installing optical bers in the position in which specular reflection light can be received and the position in which only scattered light can be received, as shown in Fig. 8. What were the background and process that led to such ideas?Answer (Saburo Okada)When the authors were engaging in the research aiming to develop a noncontact shape measurement device as a new industrial measurement system utilizing characteristics of semiconductor lasers at the time, we got a request from a local steel sheet manufacturer on whether it was possible to realize a method and technology that could detect micro-defects of micron order on the surface of high-grade rolled steel sheets, and could also identify the difference between defects and roll marks. Therefore, we decided to develop an inspection system using laser beams. The issue at the beginning was the development of a sensor that allowed quick and efficient measurement of two-dimensional distribution of laser reflection light. At the time, it was difficult to obtain a special sensor that used concentric photodiodes developed in the USA. Instead, we devised an inexpensive sensor that we made ourselves. It was a multi-split planar detector made by bundling optical bers, as shown in Fig. 6, and the objective was fullled.During the same period, an automobile parts company requested the development of an inspection system for parts with mirror treatment on the interior of cylinders with a diameter of 25 mm and a depth of 150 mm. Using the reection light image of Fig. 7 as a hint, we came up with a device shown in Fig. 8. Moreover, by increasing the focus depth by focusing the light gradually by long-focus lens after putting 200 mm or more of distance between the laser source and the flaw surface, and by optimizing the thickness, number, and tip position of the optical bers, and after much trial-and-error, we were able to efciently separate specular reection light, scattered light, and diffracted light, and were able to greatly improve inspection performance.Comment (Keiichi Ikegami)The principle of this device has been mostly covered in Chapter 3, but I think the technological highlight of this paper is presented in Figs. 12 and 13. I think it will be more useful to the readers if you provide more detailed explanation featuring these gures.Answer (Saburo Okada)I added detailed explanations to Figs. 12 and 13.Question (Keiichi Ikegami)It seems that by reducing the diameter of the probe, the geometric condition for separating specular reflection light and diffracted (scattered) light became stricter. How did you work around this? If you could explain to the extent you are allowed, I think it will be easier to understand.Answer (Yasufumi Esaki)As a countermeasure against the reduction of data types, we made it possible to nely set the judgment conditions by adding algorithms. Every year, we improve the percentage of correct answers by adding 25 to 30 functions. At the same time, we also worked on the improvement of reception efciency of laser irradiation and reflection light, to improve the quality of the obtained data.4 Key to overcoming crisesQuestion (Ken’ichi Fujii)You write that the collaborative research unit was established due to the re-organization of AIST in 2003, and that the potential demand survey was conducted during the recession period after the Lehman Shock to gain understanding of the potential demands such as for hole diameters of 20 nm or less and further time reduction, and these were greatly useful in the later product realization. You succeeded in downsizing the probe by developing a hollow motor as new transfer technology during the time, and this led to good sales performance after 2010 by increasing the degree of completion of the product. I imagine that there was much difculty in starting new development during the economic recession. What were the way of thinking and policy that became the key?Answer (Yasufumi Esaki)During recession, the most important thing, I think, is how to prepare for the time when the economy improves. There was plenty of time due to slow business during recession, and we already knew the business potential of this inspection system due to our own surveys and projections. Therefore, I was thinking about how to push the business forward in the minimum time possible. First, we narrowed down the target, and set aim on the critical parts and important safety parts of automobiles. That is because we expected that the car companies would spend a lot of budget on automated inspection for parts subject to 100 % inspection in which there was no tolerance for oversight.