Big Data Computation for Workshop-Based Planning Support

Involving stakeholders in decision-making at planning workshops requires a computer system to respond quickly to various scenarios proposed by participants. Due to the increasing complexity of urban systems, such planning support often faces the challenges of "big data" and poor computational performance. This paper proposes a new approach using parallel processing techniques (i.e. MPI Message Passing Interface) to support workshop participants in interactively building planning scenarios and visualizing outputs of job accessibility across the Greater Manchester. MPI-based parallel algorithms have been run on a cluster of computers for reducing computational time cost. The tested results and computational performances are critically evaluated and recommendations for future work provided. Particularly this paper also addresses several key research questions related to a theoretical framework of applying big data analytics for urban planning support.