Kansei, surfaces and perception engineering

The combined sensation of a products' surface gloss, colour, haptic properties like 'friction', 'elasticity', 'hardness' and 'temperature' create an intended message to the customer received as stimuli (R) by the human five senses, transformed to psychological sensation (S). Psychological sensation (S) was expressed in Fechner's law as

$$\begin{eqnarray}&&S=k\,{\rm{l}}{\rm{o}}{\rm{g}}\,R\end{eqnarray} \tag{ 1 }$$

where k is a constant and the sensation S follows a logarithmic function where small differences in stimuli create a larger variation of sensation than for changes of stimuli at higher values [3].

Later S S Stevens at Harvard developed a similar model—Stevens' power law—sensitive to the fact that different types of stimuli follow different curve shapes to psychological sensation:

$$\begin{eqnarray}&&S={\rm{a}}{I}^{b},\end{eqnarray} \tag{ 2 }$$

where a is a constant, b is a stimuli exponent varying with the type of stimulation (visual, haptic, smell, taste, or audio) and I is the stimulus energy related to stimuli (R), Fechener's law in equation (1) above [4]. So to convey a 'message' strong enough to the customer, thresholds for the lowest detection level of changes in stimuli and the function relating the stimuli to psychological sensation are important to understand.

Questions that need to be answered related to surface engineering are the minimum roughness of a handle the customer can sense and the differences of texture roughness allowing a handle with two textured parts to be perceived as having the same haptic roughness sensation, i.e. defining thresholds for texture sensation and tolerance in relation to customer expectations and satisfaction.

Aesthetics can be explained as the human perception of beauty, including sight, sound, smell, touch, taste and movement—not just visual appeal. Hidden factors controlling appreciation of beauty have been discussed by philosophers since the Ancients and were established as a subject of science when Osgood et al introduced the semantical differential method used to quantify people's perception of a product [5]. Here, a semantic scale composed of polar opposite adjective pairs separated by a five to seven point rating scale is used. For example, a customer could rate the attitude to a product by grading adjective pairs (rough to smooth, cold to warm, dark to bright) on seven grade scales. Semantic scales could then be evaluated using e.g. principal component analysis (PCA), to draw general conclusions of attitudes.

However, one important component affecting the aesthetics and attitude to the products is the customers' need or motivation. Motivation and need of a customer were discussed by Maslow [6]. Here, five levels of motivation (biological and psychological needs, safety needs, belonging and love needs, esteem needs and self actualization) were accentuated. If the psychological sensation (S) triggered by the physical stimuli matches the customers' expectation at the present motivation 'level', the attitude to the product would be positive, i.e. the message of the product design would create a positive perception of the product and its properties to the specific customer, including both the psychological sensation e.g. from surface roughness level and the customers' current 'level' of motivation from a position of wanting to satisfy basic biological needs, e.g. having simple eye glass frames without requirements of expensive, exclusive and luxury signalling high gloss polished roughness.

Schutte [7] added to the discussion of needs of the customer the pleasures of motivation by Jordan's 'four pleasures' [8]: physio—to do with the body and the senses; psycho—to do with the mind and the emotions; socio—to do with relationships and status; and Ideo—to do with tastes and values. Jordan's four groups complete Maslow's five steps and their fulfilment at the different motivation levels is of importance when designing customer motivation into the product.

Our motivation for and how we perceive a product are strongly linked to the customers' 'buy' decision. Industrial design methodology aims at creating this motivation and pleasurable product experience including meaning and message for the customer [9–11].

The aesthetic and pleasing properties of a product are of major importance in order to create motivation, interest, meaning and relevance of a specific product for the customer. Since there almost always exist alternative competitive products that fulfil basic required functionalities, the intended design of product properties towards increased customer motivation is one way of 'making a difference' and standing out from the competition. The 'equalizer' introduced by Bergman et al [12] in figure 1 below, is a tool to visualize the relative importance of design element properties (form, material, colour and surface); how they are used and tuned for a given product to create the intended motivation, meaning and message, i.e. the aesthetics and core values, intended by the industrial designer.

Zoom In Zoom Out Reset image size

In the example from figure 1 above, the 13 core values are adjectives decided by the industrial designers to define the product message of a roof-mounted bicycle carrier for the automotive industry [13]. The design element surface is decided to have its highest importance on 'user friendliness', 'aesthetics', 'well-thought out', 'quality', 'prestige' and 'professional', and consequently surface properties like gloss, average roughness and texture on the final product needed to be verified towards those core values for a successful product.

Ideaesthesia can be defined as the phenomenon in which activations of concepts results in a perception-like experience [14] To objectively and transparently judge and measure how the specification of physical design elements create the expected subjective customer perception, i.e. creating ideaesthesia, is a complex task involving both physical metrology and perceptual evaluations. An example of ideaesthesia is the experiment made by the psychologist Wolfgang Köhler in 1929 [14], showing the strong correlation between the visual shape of an object and the speech sound (see below figure 2, top and middle) named the 'bouba/kiki' effect. The word lumumba is normally connected to the top and middle right 'soft-large radii contour shape' image and the word takete with the top and middle left 'sharp angle, straight line contour shape' image. Today, a strong belief in the industrial designers' expertise and intuitive ability to make judgements exists and is regarded as 'tacit' knowledge based on skills, ideas and experiences hard to formalize for an organization.

Zoom In Zoom Out Reset image size

A tool used frequently within the discipline of industrial design and strongly related to ideaestathesia, to explain and formalize former tacit knowledge is 'semantics', the study of meaning and the relation between signifiers, like words and symbols, and what they stand for.

An example of industrial design semantics connecting design elements to core values, adjectives, is the 'softening' of the perceived visual sensation in figure 2 from a sheet metal surface by mimicking a 'soft' natural hair texture (bottom left) with the Angel Hair™ texturing³ (bottom middle), compared to the more traditional 'hard' 'Taketeish' brushed steel texture (bottom right).

The bouba/kiki effect, which later became the lumumba/takete effect, can be explained as a case of ideasthesia [15], In the example above, the design elements form and surface were possible to correlate to both the sensoric visual and sound experience of the adjective pair 'takete' (hard) and 'lumumba' (soft).

Perceptions involve any or all of the five senses. Understanding the structure of how this works can create a more robust and controlled process when industrial designers create new concepts for a predicted user experience.

The framework of perceptual product experience (PPE framework) [9, 16] considers perceptual product experience as composed of three core modes: the sensorial mode including perceptions of stimuli experienced with any of the receiver senses; the cognitive mode, where we understand, organize, and interpret and make sense of what we perceive; and finally the affective mode concerns itself with experiences that are affective: feelings, emotions and mood states, as a result of product perceptions.

The PPE model in figure 3 illustrates a model for the intended product communication between the producer and the consumer, i.e. how the industrial designers' intended product message, semantics, is expressed as core values, adjectives and converted into design elements with controlled properties creating consumer sensations, and ideally results in ideaesthesia, a pleasurable experience of the product at the customers motivation level.

Zoom In Zoom Out Reset image size

Soft metrology is defined as ''the set of techniques and models that allow the objective quantification of certain properties of perception, in the domain of all five senses.'' [17]

Soft metrology addresses a broad range of measurement outside of the traditional field of physical metrology [17]:

• psychometric measurement or perceived feeling (colour, taste, odour, touch),
• qualitative measurements (perceived quality, satisfaction, comfort, usability),
• econometrics and market research (image, stock exchange notation), sociometry (audience and opinion),
• measurements related to the human sciences: biometrics, typology, behaviour and intelligence.

The ideal would be to perform physical measurement using sensors applied to a subject placed in a test situation and thus to establish useful measurement scales correlating human subjective responses and physical objective metrology, i.e. combining traditional physical 'hard metrology' (geometry, colour, gloss, taste, smell, nois and tactile properties) to enable increased understanding of the influence of physical product properties on human responses and perception, see figure 4.

Zoom In Zoom Out Reset image size

Here, the human would be considered as a measurement system defining sensitivity, repeatability and reproducibility, and comparing the results with those obtained by methods from traditional 'hard' physical metrology.

The notion of subjectivism can of course be discussed further, related to figure 4. Parts of what are described as subjective human responses in the figure can actually be described as general perception, though subjective. For instance, the bouba/kiki effect, in which subjective perceptions are shared by all respondents, and therefore can be seen as a general perception, and not notified as a personal opinion of what is perceived.

The area of soft metrology has got a lot of attention and departments were formed both at the standards institutes NIST in the US and NPL in the UK [17–19]; and a European project—Measuring the Impossible (MINET) 2007–2010—with 22 partners from Europe and Israel including industries and academia as well as the national standards institutes NPL, UK, and SP, Sweden [20]. In addition, in 2013 L Rossi published her doctoral thesis 'Principle of soft metrology and measurement procedures in humans' stating the importance of the field [21].

Appearance is, according to the American Society for Testing and Materials (ASTM) [22], defined as 'the aspect of visual perception by which objects are recognized'.

The visual appearance of an object is a result of the interaction between the object and the light falling upon it. Colour appearance is a result of the light reflection and adsorption by the pigments. Gloss is created by the reflection of light from the surface, and translucency is a result of light scattering while the light passes through the object (figure 5). The described complexity of the object's appearance causes different measurement technologies and instruments to be employed when attempting to quantify it [17]. Texture is a complementary component of the visual appearance and also needs to be considered.

Zoom In Zoom Out Reset image size

The concept of total appearance has been introduced to extend the concept of the appearance of an object. The total appearance, however, would require a description of the shape, size, texture, gloss and any other object properties possible to detect by our five senses (visual, haptic, smell, sound and taste) and interpreted by the brain as a 'total appearance' of an object [17, 23].

The total appearance (figure 6) could also be described as a combination of three aspects of appearance:

Zoom In Zoom Out Reset image size

Physical: object properties physically detectable by our senses, modified by the surrounding, properties of the illumination, individual factors like ageing, handicap, etc, affecting our sensibility.

Physiological: the neural effect when human receptors are subjected to physical stimuli and convey signals to the cerebral cortex, creating a sensation

Psychological: created when sensations are interpreted by the cortex, recognized as an object, and combined with inherited and taught response modifiers (memory, culture, fashion, preferences). Figure 8 summarizes the factors affecting the total appearance, resulting in two appearance images: the impact image and the sensory image. The impact image is the initial recognition of the object or scene (the gestalt), plus an initial opinion or judgement. For the sensory appearance image, three viewpoints are used to create the total appearance: sensory, emotional and intellectual. The sensory viewpoint describes thoughts associated with the colours, gloss, etc, of the object. The emotional viewpoint associates the emotions connected with colours, gloss, etc, while the intellectual viewpoint covers other aspects associated to the object and situation rather than sensory or emotional associations [24, 25].

Total appearance is closely related to the models of Intended product communication and the PPE framework, and could be used when quantifying customer perception and satisfaction using soft metrology to correlate physical and human factors contributing to product appearance images.

After Osgood's publications [5] more methodologies, e.g. quality function deployment (QFD) and the Kano model were developed with similar motivation [26, 27]. Those methods are very capable of dealing with psychological sensation but not as capable when it comes to translating the subjective sensation into design parameters, i.e. real product features influencing the perceived sensation.

Nagamachi developed the method 'affective' or 'Kansei' engineering (KE) in the 1970s which has its roots in the Japanese concept of Kansei: 'intuitive mental action of the person who feels some sort of impression from an external stimulus' [28–31]. By using the framework of KE as an approach and focusing on finding correlations between the functions, customer requirements, function requirements, design requirements and process requirements, a higher level of user quality and a methodology for soft metrology as discussed above could be obtained.

According to Nagamachi, the Kansei concept includes 'a feeling about a certain something that likely will improve one's quality of life'. KE can also be defined as a customer-oriented approach to product development. The basic idea is that the client's feelings shall be observed at the phase of idea generation in the product development process, which then facilitates the project later on when a concept reaches the production stage.

Kansei engineering handles six different phases/steps [12, 28–31], starting with the definition of the products' domain and context, see figure 7.

Zoom In Zoom Out Reset image size

The six phases range from a pilot study where the product or service is defined, including specification of the product and market, to synthesis and modelling the result of the given study.

• 1.  
  Pilot study: defining the product domain, market and users, i.e. what, who, where, why, when and how; e.g. Interior surface textures for small urban four-wheel drive cars.
• 2.  
  Describe the experience (span the semantic space): collect adequate adjectives, expected by the customer from the domain to define the physiological requirements, e.g. organic, innovative, stimulating, robust.
• 3.  
  Define key product properties (span the space of properties): in this phase it is important to find physical product properties that affect the users' view of the total appearance and experience, e.g. surface texture anisotropy, surface texture amplitude and number of cavities in a surface.
• 4.  
  Connect the experience to product properties: by using qualitative and/or quantitative studies on focus groups, connections between customer adjectives (phase 2) and design elements (phase 3) can be made, e.g. organic surfaces possess high texture amplitude and isotropic texture patterns.
• 5.  
  Validity check point: when the qualitative correlation in step 4 is established, it is important to verify the results quantitatively by verifying experiments or virtual simulations, e.g. physical tests on surfaces passing the acceptance level of cleanability in step 4 all possess a texture mean amplitude less than 0.8 μm.
• 6.  
  Modelling the domain: design and validation of a 'prediction model', e.g. surfaces passing the acceptance level of cleanability and hygiene in step 4 all possess a texture mean amplitude lesser than 0.8 μm and the cleanability increases linearly for mean amplitudes from 0.8 μm to 0.03 μm in the domain of steam sterilizers for medical applications.

A full project ranging over all the six phases of KE will result in a model or prediction of the total appearance or a limited sector of the total appearance, e.g. visual appearance or total appearance of the texture of a product within the domain selected in phase 1 of the project. Visual appearance is limited to optical factors of the product appearance while the total appearance of the texture of a product includes visual, haptic and other sensory aspects of the total appearance of the micro- or nanometre texture experience.

Below in the following sections, the Kansei methodology briefly described generally in the sections above will be detailed and exemplified with results from current and past cases and studies to illustrate the potential of an application of affective engineering concepts on product surfaces.

In this phase, it is important to define product domain and users; define and analyze what, who, where, why, when and how. In the pilot study, see figure 8, there are a lot of questions to be answered and it is crucial to navigate efficiently in the right direction from the beginning in a development project. Product designers in general have good internal navigation and relation to the design process, however sometimes the lack of communication in a project can result in endless discussions of what the aim is for whom. The 'Design Compass' and persona studies as well as mood boards are useful Kansei tools and work as external stimuli and guidelines in the process of affective engineering to facilitate the workflow in order to focus on the primary questions (what, who, why, where, when and how) in the pilot study.

Zoom In Zoom Out Reset image size

In this phase it is important to find psychological emotions and expected total appearance and perception related to a product expressed as adjectives, 'Kansei words', and grouped in logical clusters (see figure 9).

Zoom In Zoom Out Reset image size

The idea of describing the product experience using adjectives is about framing the emotional functions, i.e. defining the expected perception and total appearance. To be able to do that there is a need to 'span the semantic space', collect the expressed 'emotions', by collecting adequate describing words, which the user expresses when interacting within the product domain. By using describing words it is possible to find appropriate expressions and clusters of Kansei words expected to be associated with (and not!) a product designed to evoke an intended perception or ideaesthesia.

Today, the collection phase is optimized to be able to offer a more cost-efficient process and service with a higher quality. The 'word game' [12], see figure 10, was developed to serve as a physical tool for designers to be able to set high quality core values for a product or service in a structural way, and faster than before. Instead of implementing questionnaires, a focus group is participating in a physical word game.

Zoom In Zoom Out Reset image size

The 'word game' is implemented, sorting out ambiguous words which do not fit in the domain and context, and grouping words which are considered as synonyms among the describing words, or words that could be directly connected to each other, e.g. the words 'organic' and 'natural' could be clustered with each other even though they are not synonyms. The amount of words is reduced during this part of the project and clusters of belonging words are created. Figure 10 illustrates the last implementation step of this phase and an example of word clusters from a study of building exteriors representing the core values and Kansei words for the building exteriors [32].

In this phase it is important to find physical product properties that affect the user; analyze the properties of the domain; define properties that affect; and isolate significant properties.

When the identification of the core values and Kansei words is made, the next step is to identify properties of the existing product that can be controlled and affect the product towards those core values, i.e. design elements. In product design the primary design elements are form (geometry/shape), colour (hue, saturation, whiteness and blackness), material (chemical substance or raw material, isotropic or anisotropic, structure and strength) and surface (texture ratio, fractal dimension and directionality, texture amplitude, texture spatial properties and texture geometrical feature), in accordance with ISO 25178-2:2012, see figure 11.

Zoom In Zoom Out Reset image size

The design elements should be appropriately measurable using standardized methods and parameters like the surface texture field, stratified and feature parameters in accordance with acknowledged ISO 25178 series of standards.

Now, the correlations between the experience and 'feeling' (psychological requirements) and the functional requirements (physical requirements) have to be established as well. For instance, the adjectives clean and hygienic, expressing customer psychological requirements for a surface in a medical environment, are connected to demands on: cleanability thus related to chemical resistance against stains and cleaning agents; and scratch proofness to withstand negative wear and cleaning effects on the surface [33, 34]. ASME standard [35] connect today's requirements of hygienic surfaces to texture average arithmetic amplitude (Ra) according to ISO 4287:1997 and it is considered sufficiently smooth when the Ra value for a given surface is <0.8 μm [33, 34].

Here the ASME standard defines the surfaces' texture mean amplitude as the design element controlling the feeling and psychological requirements on clean and hygienic. Figure 12 is an example of four surface variants exhibiting different areal texture average arithmetic amplitudes possessing different levels of cleanability and hygienic properties [33, 34].

Zoom In Zoom Out Reset image size

All surfaces are well below the ASME level on the surface design element mean amplitude defining the upper threshold for a hygienic surface from 2009 but the influence of other recently standardized surface texture properties like anisotropy, fractal dimension, etc, not included in the standard could supply future more detailed information more sensitive to the requirements than the mean amplitude averaging other texture details of the surfaces in figure 12.

By using qualitative studies on focus groups, connections between Kansei words and design elements can be made. An important tool to visualize the connection between Kansei words and design elements contributing to the total appearance is the 'equalizer'.

For a given domain identified in step 1, the key product properties (design elements from phase 3) are connected to the Kansei words from phase 2, in this 4th step by, for example, using focus groups (figure 13).

Zoom In Zoom Out Reset image size

In the example below from a study made on the domain architectural external building panels [32], the valuation of the Kansei words was made by 20 people on a semantic scale 1–10 for each of the four different surface textures in figure 14 (left).

Zoom In Zoom Out Reset image size

A useful tool to visualize the correlation between surfaces' appearance with the design elements is the equalizer in figure 15, also described in section 3.2 above [12].

Zoom In Zoom Out Reset image size

The equalizer is a dynamic tool owing to the possibility to visualize the total influence and total appearance of the combination of design elements (the form, material, colour and surface) on the selected Kansei words developed in step 2 above.

When the correlation in thesynthesis in step 4 is established, it is important to verify the results by statistical tests, experiments or virtual simulations.

The validity checkpoint is about an overall validation of the concept's total appearance, verifying quantitatively the Kansei words from phase 2 and their connection to the design elements obtained in phase 4. Practical testing of concepts is made and quantitatively evaluated by the correlation of 'soft' judgements of Kansei words to 'hard measurements' of design element properties.

In affective surface engineering, a connection between Kansei words and surface texture parameters describing the micro- and nanotopography is made. In a study made by the authors on wall panels for hot bath sauna interiors [32], three test panels of 10 people with experience from general product development, specific sauna product development and general sauna bathing, graded the 11 surface textures on a seven grade scale against the eight emotional Kansei words and three design element-related surface areal texture property parameters from the ISO 25178-2:2006 (see table 1).

Table 1.  Correlation table between the sauna Kansei words and surface texture areal parameters according to ISO 25178-2:2006 (significant parameters in bold) defining three design elements associated to the surfaces. Reproduced with permission from [32].

The Sq, Str and Sdq parameters (texture average amplitude, texture ratio and texture slope) had a significant correlation with the Kansei words beautiful, stylish, and attractive. Here the soft metrology adjectives are demonstrated to correlate to 'hard' metrology measurable surface texture parameters, i.e. the example validates the principle that Kansei adjectives can be linked to ISO standardized measurable parameters, thus defining geometrical features controllable by manufacturing engineering and possibly to use to specify product requirements.

The final step is intended to create a model that combines, refines and describes the results from the previous five phases in the Kansei methodology. Hence, to assemble a model bridging the emotional semantic and product properties' space.

In the project with a sauna manufacturer, a design manual was made on request of the company. The design manual basically linked surface geometrical properties (the significant design elements and properties) to the Kansei words according to the results demonstrated in step 5 above. In practice this resulted in designer rules collected in a physical booklet for the context of sauna wall panels, establishing that:

• A low average roughness (Sa) and low surface slope (Sdq) increase the stylish and attractive emotions of a sauna wall texture,
• The same surface combinations together with increased texture anisotropy, e.g. a directional pattern, will also affect the beautifulness of the surface.

This realizes a product development tool for the sauna company to facilitate the material and surface design of their products connecting to the needs and expectations of the customer.

In many cases it is desirable to formulate models describing quantitatively the relation between design elements and the desired soft metrology Kansei words as a complement to the design manuals with qualitative designer rules. In a study [36, 37] in the context of tissue paper haptic appearance a complex model was synthesized using eight constants (A–H), three material properties (layer type (DL), stiffness, elasticity (stretch)) and four areal ISO 25178 surface texture properties (peak material volume (Vmp), core height (Sk), maximum height (amplitude, Sz), autocorrelation length (repeating wave length, Sal)).

$$\begin{eqnarray*}&&\begin{array}{c} \mbox{\textasciigrave } {\rm{p}}{\rm{e}}{\rm{r}}{\rm{c}}{\rm{e}}{\rm{i}}{\rm{v}}{\rm{e}}{\rm{d}}\,{\rm{h}}{\rm{a}}{\rm{p}}{\rm{t}}{\rm{i}}{\rm{c}}\,{\rm{r}}{\rm{o}}{\rm{u}}{\rm{g}}{\rm{h}}{\rm{n}}{\rm{e}}{\rm{s}}{\rm{s}}\mbox{'}=A\,-\,B* DL\\ +{\rm{C}}* Vmp+D* Sal+{\rm{E}}* Sk\\ +{\rm{F}}* {\rm{s}}{\rm{t}}{\rm{i}}{\rm{f}}{\rm{f}}{\rm{n}}{\rm{e}}{\rm{s}}{\rm{s}}\,-\,G* {\rm{s}}{\rm{t}}{\rm{r}}{\rm{e}}{\rm{t}}{\rm{c}}{\rm{h}}\,-\,H* Sz.\end{array}\end{eqnarray*}$$

In the equation above, the product design properties DL, Sz, and stretch have a negative regression coefficient sign, showing that an increase in the parameter value results in a decrease in 'perceived haptic roughness'. The coefficients for Vmp, Sk and Sal were positive, hence positively correlated with increased (improved) perceived haptic 'roughness', i.e. increased texture peak material volume, core height, autocorrelation length and decreased maximum texture height improve the customers' haptic perception of tissue products within the context of the performed study.

The designer rules and the equation above are examples where affective engineering and soft metrology results are synthesized into tools able to predict customer perception and aspects of total experience supporting organizations' possibilities to maintain customer focus and competitive advantage.