Quality assurance in the evaluation of virtual prototypes by using eye tracking to optimize the cus-tomer requirements analysis (ReqET)

• Applicants: Professor Dr. Thies Pfeiffer; Professor Dr.-Ing. Robert Refflinghaus
• Subject Area: Human Factors, Ergonomics, Human-Machine Systems ; Image and Language Processing, Computer Graphics and Visualisation, Human Computer Interaction, Ubiquitous and Wearable Computing ; Data Management, Data-Intensive Systems, Computer Science Methods in Business Informatics ; Engineering Design, Machine Elements, Product Development
• Term: since 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 515614446

The increasing integration of the costumer into the development process (Customer-Co-Creation) is accompanied by the conviction that the development of high-quality products is only possible if product features meet customer requirements as comprehensively as possible. Design reviews are therefore used to validate customer requirements as early as possible. However, the user studies used in this process are extremely complex, while their results are subjective. Iterative or agile requirements validations are therefore currently hardly economically feasible in hardware-intensive consumer products. The goal of the research project is therefore to carry out customer requirement analyses on virtual prototypes and to increase the precision, comprehensiveness and reliability of their results compared to classical approaches by using eye-tracking. Through (semi-)automated evaluations, customer requirement analyses can thus be objectified and significantly accelerated. The quality of survey methods is assured via automated behavioral observations during viewing and interaction with prototypes. To achieve this aim, an integrative technical framework will first be developed in which virtual prototyping is combined with eye-tracking, and attentional patterns can be systematically recorded and analyzed on interactive three-dimensional machine models. Based on this, a model of the interdependence between the tracked behavioral data (perception level) and the statements about requirements on the decision level will be successively created and validated through studies on concrete product contexts. Overall, it will be researched whether valences and prioritizations of decision processes can already be extracted from the objectively collected perception data, and thus faster, with sufficient precision, and to what extent customer requirement analyses can be enriched and made more accurate through synergy effects of semantic-visual fusion. For this purpose, three approaches are investigated and further developed: The Visual-Semantic-Anchoring-Approach, the Interaction-Intensity-Resilience-Approach and the Revealing-The-Unexpected-Approach. The influencing factors to be examined more precisely are the connections between the visual examination of product features and the reliability and relevance of statements on these features; the concretization of linguistically under-determined localized statements through the direction of gaze, as well as the discovery of previously subconscious or non-verbalized requirements through unexpected (gaze) behavior patterns. These approaches form the basis for the development of a holistic process model for the interaction between behavioral data and (expressions of) requirements, which is iteratively refined and validated. The finalized model eventually pro-vides the basis for eliciting high-quality customer requirements and thus ultimately leads to products of higher quality.

• DFG Programme: Research Grants