||This study is to find out the Critical Needs of the common service design of Ikari coffee external customers and internal customers, and combine the seven major expression recognition researches of customers with ambient light and music. With the ambient light color changing from cold to warm in 9 colors and 4 different styles of music, we explore the changes in the expressions of the guests. The purpose is to enhance the joyful expressions of the guests; every day is a wonderful day.
Research design 1: First we explored the critical demands of external customers for service design. We investigated customers’ basic data, including purchasing frequency, occupation, age, sex, and marital status before interviewing 55 customers face to face, 27 of whom came from franchise stores and 28 of whom came from regular chain stores. We roughly inquired why they came to our stores and then interviewed them according to 4 domains: interpersonal interaction design, space design, operating procedure design and ideal-café design. After collecting the data of the interviews, we concluded the critical demands of external customers by categorizing and inductive method.
Research design 2: Then, we explored the critical demands of internal customers for service design. The design is based on the 70 subjects consisting of the frontline supervisors, coworkers on duty and team leaders. We adopted brainstorming, KJ conception and cause and effect diagram, mind-mapping to conclude 6 facets of critical demands for service design from interior customers. The six facets are career development design, environmental design, operating procedure design, working design, brand-marketing design and trans-department cooperation design. Top 5 service designs requiring improvement most in every facet are found. Thus, the detailed items are amounted to 30 and we draft an operation scheme.
Research design 3 is to perform the experiment by combing external customers and interior customers’ mutual critical demands and lights as well as music. We detect expressions through ambient light colors and musical changes and further recognize and analyze customers’ emotions. We build the ESTD（Expressions State Transition Diagram）and choose Philip hue API smart ambient light and music, using Ada- Boost detects the face and ASM（Active Shape Models） to capture the features of the facial features, then calculates the movement of the facial features and feature points. After the calculation is completed, the ANN (Artificial Neural Network) is used to classify the expressions to generate the identification results, the environment of the control group is with fluorescent light but without music and that of the experiment group is with the smart ambient light, with music and the smart ambient light with music.
The test performed a total of 1,008 actual expressions. Under the influence of external environmental factors, the number of successful identifications was 783, 342 females and 441 males. The successful recognition rate was 77.7%.
(一) In the joyful expression analysis of the expression recognition,1.The number of times of showing joyful in experimental groups increased by 4.8%. 2. (1) In the atmosphere of smart-ambient light analysis, the joyful expressions of females increased from 23.8% to 31.25%. (2) In the atmosphere of music, the proportion of females to all females increased from 13.8% to 23.3%. The proportion of males to all males increased from 9.1% to 16.0 %. (3) In the smart-ambient light vibe, however, when playing music, the proportion of joyful expression of females to all females largely increased from 1.3% to 36%. The ratio of females to total increased from 0.6% to 9.6%. 3. The positive expression "joyful” at t test (n < 30) was significant (p < 0.05).
(二) Positive expression analysis, 1. The experimental group and the control combination accounted for an increase in the positive expression and the expression of the joyful expression, reaching 100% with the purpose of our experiment. 2. After receiving the stimulation of “ambient light", "the negative expression of women's negative expression decreased significantly" (p < 0.05). 3. Under the stimulation of "music", the subject's "positive expression is significant" (p < 0.05). 4. In "music and ambient light", the "positive expression is significant" (p < 0.05).
(三) Percentage analysis of facial expressions analysis of joyful expression, verifying that 18 items are assumed to be established, of which 11 items are established, accounting for 61%, and 7 items are not established, accounting for 39%.
The design of workshops and the application of enterprises for the experiment in coffee shops are consequent attempts. All-dimensional expression change recognition can not only detect the people’s emotions but include their feeling inside as a breakthrough. The ultimate goal is to create a happy coffee shop.
Keywords: service design, key requirements, ambient light, expression recognition.