Scenario analysis is a fundamental practice in urban planning in which a decision-maker formulates a question and a computer model simulates the hypothetical scenario for subsequent assessment. In this talk, I concentrate on visual scenario analysis for sizing and rebalancing tradeoffs in mobility on demand (MoD) systems. Developing such models is challenging because trip patterns are by definition random while rebalancing patterns require first solving routing, a computationally intractable problem even for medium-sized MoD systems. Working with data from bike sharing systems in Boston, NYC, and Washington DC, I present a novel visualization approach based on reconstructing accumulation dynamics, showing that, independently of system size, network topology, or trip pattern characteristics, vehicle mass in MoD systems moves periodically between two types of locations and accumulates between two other types of locations. Next, I present a method to build data-driven system dynamics macromodels that allow planners and decision makers to interactively and visually explore scenarios like: by how much will traffic, fleet, or parking requirements change for a marginal change in rebalancing work? I show that land use patterns and basic operational decisions, may define performance limits for MoD systems that no technology can overcome.
Dimitris Papanikolaou is an Assistant Professor at UNC Charlotte, jointly between the School of Architecture and the Department of Software and Information Systems. His research investigates how the study of system dynamics, social behavior, and information technology can inform design strategies for intelligent urban, building, and mobility systems. His interests include data visualization, ubiquitous computing, human-computer interaction, computer supported cooperative work, complex systems modeling and simulation, urban operations research, and design research. He has worked at Microsoft Research on applications of the internet of things and at the MIT Media Lab co-developing Mobility on Demand, an intelligent sharing system of electric foldable cars named by TIME magazine as the best automotive invention of 2007. Dr. Papanikolaou holds a Doctor of Design (DDes) from Harvard GSD, an MSc from MIT Media Lab, a SMArchS in Design Computation from MIT Architecture, and a Diploma in Architectural Engineering from NTUA