Towards Formalizing Cyber-Empathic Design—A Data Driven Framework for Product Design
Ghosh, Dipanjan Dipak
MetadataShow full item record
A critical task in product design is mapping information from consumer to design space. Currently, this process largely depends on designers identifying and mapping psychological and consumer level factors to engineered attributes. In this way, current methodologies lack the provision to test a designer's cognitive reasoning and could introduce unnecessary bias during this mapping process. This dissertation develops novel data-driven customizable frameworks for product design, which map user perceptions using latent variables and product usage contexts, identified with product embedded sensors, to consumers' product attribute preferences. In the first section of the dissertation, the foundational framework for this research, Cyber-Empathic (CE) design, is developed to integrate user-product interaction with psychological latent factors to model user perceptions. The framework leverages concepts from Structural Equation Modeling (SEM) and unsupervised machine learning methods. The effectiveness of the CE framework is validated using three sensor-integrated shoe cases studies. The second section of the dissertation extends the CE design framework to incorporate product usage contexts identified using user-product interaction data. Machine learning architectures based on Convolutional Neural Networks and Generative Adversarial Networks are developed to identify usage contexts. The effectiveness of the architectures to identify product usage context are demonstrated using two case studies involving activity recognition. Finally, the third section of the dissertation, extends the CE design framework to model user preferences and the causal structure of those preferences. An extended framework, the Discrete Choice Analysis based Cyber-Empathic (DCA-CE) design framework, is developed which integrates choice modeling, CE design and usage context. Using a simulated case study of sensor-integrated shoe design, the effectiveness of the DCA-CE framework is validated. The frameworks developed in this dissertation represent innovative approaches to leverage user-product interaction data to model user perceptions and usage contexts, assisting designers in making informed product design decisions by uniquely understanding the causal structure of preferences.
Showing items related by title, author, creator and subject.
Seismic Resistance of Reinforced Concrete Frame Structures Designed Only for Gravity Loads - Part I: Design and Properties of a 1/3 Scale Model Structure Bracci, J.M.; Reinhorn, A.M.; Mander, J.B. (1992)This report is Part I of a three-part series prepared for a comprehensive evaluation of typical gravity load designed low-rise reinforced concrete frame buildings. Only the general non-seismic detailing provisions of ...
Kemper Lewis Principal Investigator (2014-04-02)The University of Buffalo-SUNY (UB) is planning to join the Industry/University Cooperative Research Center (I/UCRC) entitled "Center for e-Design and Realization of Engineered Products and Systems (e-Design)" ...