Product Service System Design in New Situations: Prediction of Demand Surfaces From Environment

Abstract

Product service systems (PSS), such as DVD rental stations or the subway, face a unique problem slowing their adoption and growth: they are uniquely dependent upon timely or expensive user data for system planning, yet user datasets are only accurate for a small part of the entire PSS. Thus, methods to use the available data effectively and use data collected in one portion of a PSS for system design in another portion could transform PSS design. PSS allow customers to purchase use of a product rather than the product itself, resulting in improved environmental sustainability. The central question examined by this work is: how can designers compensate for situations where the design environment has changed and limited user data is available to inform demand estimations? Our hypothesis is that publicly available socio-demographic and environmental variables can be used to estimate the demand outside of the boundaries previously constrained by available user data. This approach was validated by applying multivariable regressions to a major Bike Share System (BSS) Expansion, outperforming the methods utilized by the BSS operators. The approach is tested in four different design scenarios. When examining all 174 stations added in 2015, our approach shows a moderate correlation with the ideal ordering (Rho = .566, Stations = 174, p < .01), while the implemented operator ordering was only weakly correlated (Rho = .334, Stations = 174, p < .01). This work demonstrates a partial solution to the problem of transforming available user data into demand for new situations.