• Journal of Intelligence and Information Systems
• /
• v.28 no.1
• /
• pp.287-310
• /
• 2022

Companies' interest in developing AI-based intelligent new products is increasing. Recently, the main concern of companies is to innovate customer experience and create new values by developing new products through the effective use of Artificial intelligence technology. However, due to the nature of products based on radical technologies such as artificial intelligence, intelligent products differ from existing products and development methods, so it is clear that there is a limitation to applying the existing development methodology as it is. This study proposes a new research method based on KANO-TOPSIS for the successful development of AI-based intelligent new products by using car voice assistants as an example. Using the KANO model, select and evaluate functions that customers think are necessary for new products, and use the TOPSIS method to derives priorities by finding the importance of functions that customers need. For the analysis, major categories such as vehicle condition check and function control elements, driving-related elements, characteristics of voice assistant itself, infotainment elements, and daily life support elements were selected and customer demand attributes were subdivided. As a result of the analysis, high recognition accuracy should be considered as a top priority in the development of car voice assistants. Infotainment elements that provide customized content based on driver's biometric information and usage habits showed lower priorities than expected, while functions related to driver safety such as vehicle condition notification, driving assistance, and security, also showed as the functions that should be developed preferentially. This study is meaningful in that it presented a new product development methodology suitable for the characteristics of AI-based intelligent new products with innovative characteristics through an excellent model combining KANO and TOPSIS.

• Journal of the Korean Association of Oral and Maxillofacial Surgeons
• /
• v.27 no.3
• /
• pp.214-220
• /
• 2001

The clinical application of the three-dimensional radiographic technique had been limited to standard Broadbent-Bolton cephalometer with biplanar stereoradiography. We developed a new method for compensating the error of head position in ordinary non-biplanar cephalostat. It became to possible to use the three dimensional cephalogram commonly in clinical bases. 1. The method of methemetical compensation of head positioning error in non-biplanar condition was evaluated with dry skull. The error of the method of first and the second trial was $0.46{\pm}1.21$, $0.33{\pm}0.90mm$, which means the error of the head positioning correction in conventional cephalogram was within clinical acceptance. 2. The reproducibility of this system for clinical application was 0.54 mm ($-2.99{\sim}2.26mm$) which defines the absolute mean difference of the first and second trial. Compare to the The landmark identification error $1.2{\pm}1.6mm$, the error of the measurement was within the range of landmark identification error. The result indicates the adequate clinical accuracy of the computation of three-dimensional coordinates by compensation of the error of the head position in ordinary non-biplanar cephalostat.