Design Science PhD candidate Alex Burnap is taking crowdsourcing to the next level. As part of an interdisciplinary team spanning mechanical engineering, statistics, psychology, and computer science. Alex is studying methods of crowdsourced evaluation for design processes – and their limitations – in order to better understand end-user preferences.
The National Science Foundation (NSF)-funded project, entitled Creativity through Collaborative Human-Machine Interactions: A Formal Approach to Design Crowdsourcing, examines methods of crowdsourced evaluation that push the boundaries of scale and interactivity, allowing for key findings on data aggregation in relation to customer insight and design innovation. Among their important contributions, Alex and his team are the first to use modern machine learning techniques, particularly “feature learning methods,” on design and marketing data – specifically, modeling the heterogeneity of customers in a crowd in order to predict design preferences for large automotive companies. They have also utilized 2D/3D realistic morphable models using real-time GPU computation, enabling more interactive crowdsourcing.
How will this research affect real designers in their everyday work? As Alex puts it, “These are a set of powerful tools that might be additional tools within the designer’s toolbox.”