Material development based on emotions
Why do you like a material or not? If you are looking for an answer, you have to look the customer in the eye and measure and evaluate his emotions. Imat-Uve now wants to implement nothing less with a partner in the "Eliot" project.
The entire industry is currently following this credo by placing the customer even more at the center of all development considerations in the future. In addition to the focus on costs, it is the customer who makes the most emotional decisions. And especially when it comes to the interior, emotionally guided perception is a decisive component.
Established market research attempts to depict emotions through cognitive judgments from individual or group surveys. The researchers also use behavioral observations. For Leander Schweiger, psychologist for Data Science and UX Research, and his employer Imat-Uve, these approaches represent emotions only indirectly: "A great deal of interesting information is lost that needs to be used for the development and design process," says Schweiger.
For this reason, the development and engineering company has developed a process called "Eliot" (emotional evaluation of things) for the targeted design of material and design experience. Since June 2018, in cooperation with the Neurosensoriklabor of the University of Kassel, a measuring method has been developed which, according to Imat-Uve, should enable the objective description and evaluation of the sensual material experience.
The ideal way of measurement
The partner in Kassel was chosen very consciously because neuro sensors describe objects through the various sensory channels of the human being. In this way, the researchers and engineers want to obtain a sound overall picture of the user's world of experience. Among other things, disturbing influences must be excluded. The time factor is also decisive: "A query in real-time, during the experience, is the ideal solution in principle," explains Schweiger. Eliot transforms this claim into a standardized procedure. Direct access through the automated use of electroencephalography (EEG) with the help of a so-called brain-computer interface is to represent the core of the development. Via this system, the partners collect the sensory experience of the user in a completely standardized structure. As a combination of biometrics and questioning, this system maps both verbal and non-verbal content.
What is unusual about this development project is that the project leader of the study is an employee of Imat-Uve and the university and heads the research team at both locations. "This is a pilot project that was chosen in order to ensure that the intermeshing of scientific and market interests is as effective as possible," emphasizes Schweiger. Eliot's pilot development will be carried out by the University of Kassel, while Imat-Uve will then be responsible for the development of market maturity.
The measurement method: Robotics in use
In detail, the measurement procedure is as follows: A selection of materials, surfaces, and components are stored in sample holders. A robotic system presents them to the test person one after the other during the test. It is used to make the testing process as efficient and objective as possible and is used for standardization to counteract interferences such as test leader effects. During the material presentation, the subject's EEG, pupil behavior and mimic expression are tracked. The EEG is based on the clinical standard, a depth camera is used for pupils and facial expressions. It can show the face of the test person in spatial resolution. The first presentation is followed by a spontaneous rating. In the second phase, both the physical and affective characterization of the material is carried out by means of a parallel query of the test persons and stimulus presentation.
The cooperation partners want to measure the emotions of at least 100 test persons in order to model correlations between the data types (EEG, eye tracking, survey, technical test data) using a method of machine learning. These correlations can be used to make a prediction from one data type to another. "Successful learning of the neural network can significantly shorten the testing and development process," says Schweiger.
Imat-Uve has a clear roadmap: From the end of 2019/ beginning of 2020, the first field trials are to be carried out with associated partners. Customer interest seems to be high. According to Leander Schweiger, this is why a lot of input was received from them through regular discussions: "The system was designed to be as flexible as possible and is still so open today".
This article was first published in German by Automobil Industrie.