Kansei Technology
* Kansei indicates holistically human “Feelings,” “Preferences,” and “Values.”
Throughout our long history, Yamaha has been in touch with our customers’ voices and experienced their relationships with music. The deep understanding of “feelings,” “preferences,” and “values” regarding sound and music that we have cultivated through these experiences – in other words our “knowledge of Kansei” – is a strength that clearly differentiates Yamaha from other companies. Our goal is to systematically utilize this “knowledge of Kansei” through Kansei technology. The “knowledge of Kansei” possessed by each of our employees can be visualized in language, charts, and numerical values, and shared for systematic utilization. We believe that this will further strengthen our “knowledge of Kansei” and contribute to the creation of new value unique to Yamaha.
How Kansei Technology Works
Kansei technology models human Kansei based on a simple system that “outputs” a response when a person receives a stimulus as “input.” For example, when a person hears “the sound of a musical instrument (input),” the system describes it as “beautiful (output).”
In modeling, psychometric, physiological, and physical measurements are used to convert input and output into data. Furthermore, this relationship between input and output is visualized by analyzing statistics and AI techniques. This will allow our organization to make use of the “knowledge of Kansei” possessed by individuals.
Note that when converting input and output into data, an environment that minimizes the influence of factors that are outside the scope of measurement might be required. To ensure subjective evaluation, Yamaha uses a Kansei laboratory that combines room acoustic performance, elevation grids, and lighting equipment in accordance with international standards (ITU-R BS.1116-3).
Kansei Laboratory
Commitment to Kansei Technology
Yamaha is making systematic use of its “knowledge of Kansei” to utilize Kansei technology based on the following three approaches.
Psychophysical Modeling
When developing musical instruments, we place great importance on “understanding the sound perception and preferences of players and listeners.” In many cases, the production of musical instruments is a team effort. However, if the team members have different ways of expressing their “understanding of the sound perception and preferences of instrumentalists and listeners,” the goals of the team might become unclear. To resolve this, a shared language is needed. However, there is not necessarily a shared language for some aspects of the world of musical instrument production. For example, “dark” is one of the most important words used to describe the sound of wind instruments. Since “dark” is an abstract expression, it might not naturally have a shared understanding. Therefore, in order to model the overall picture of the elements that make up “dark,” we described sound impressions in concrete terms that are clearly understood by all developers and connected them with physical features. This model allows developers to obtain a shared understanding of “dark,” understand the physical features that can be used in design, and clarify the design guidelines that the team should aim for. We believe that by doing so, we can provide instruments with a “dark” sound that is unique to Yamaha, while also catering to the perception and preferences of wind instrument players.
Clarinet sound evaluation experiment
Formalization of Artisanal Skills
Yamaha has cultivated artisanal skills on the production floor. For example, when inspecting the wood-grain of high-end interior parts for cars, we rely on the expert skills of inspectors to accurately determine quality. It was difficult to describe the expert skills of these inspectors in words that can be shared within our organization. However, if we can visualize the expert skills of inspectors and share this throughout our organization, it will further improve the efficiency of the production process and stabilize quality. To achieve these goals, we developed an AI system to visualize inspection skills. Specifically, we used a machine learning model to simultaneously learn the image data of the grain and the inspection data in order to build an AI system that visualizes not only how inspectors judge the quality of the grain, but also their criteria for doing so (grain hue, stripe shape, etc.). The judgment criteria visualized by this AI system will be used as proprietary criteria at Yamaha production sites. Inspectors can use these criteria to make more stable and precise judgments, further improving the efficiency of the production process and stabilizing quality. In addition, we can clearly provide our customers the value and reliability that only Yamaha can offer. Through the use of AI technology that further enhances artisanal skills, we will continue to provide our customers with high-quality products backed by the skills of our artisans.
Visualization of Diverse Sensitivities and Values
“Feelings,” “values,” and “preferences” in the world of sound and music vary widely, and naturally express the personality of each individual. Yamaha uses Kansei technology in a constant effort to capture this diversity from both quantitative and qualitative perspectives. We believe that enhancing this approach will lead to the creation of value that is unique to Yamaha and closer to people than ever before.
The Future of Kansei Technology
Yamaha will continue to develop measurement techniques and expand measurement targets in order to achieve further developments in Kansei technology. This will allow us to contribute to the creation of new value to support the “well-being of people around the world.”