Data Science for Processing Networks

Hospitals, courts and public transportation are service networks of central societal importance. Due to technological advances and public awareness, processes in such networks create (or could create) vast amounts of operational data, at unprecedented resolution and quality. Yet harnessing this data for publishable, reproducible and scalable research is an acknowledged and yet-to-beovercome challenge. What we offer here is a blueprint for data-science labs, which will enable transformative steps to address this challenge: specifically, bringing together multiple scientific disciplines to the study of processing networks, which will support their design, planning, control, prediction and improvement along dimensions such as operational efficiency, quality of outcomes, and fairness of access.This is therefore a proposal to develop infrastructure—physical (a data lab), human (a research team) and scientific (scholarly knowledge)—to advance data-science for processing networks in general, and healthcare systems in particular. The Processing-Network Lab, or PNLab for short, will serve as a data-based research hub which, among other things, will facilitate understanding, dissemination of and improvement upon best practices. PNLab activities will rely on analysis of its data and validation of models against this data; they will then foster the creation from data of novel models (eg statistical, mathematical, computational), scientific principles (eg congestion laws) and engineering solutions (eg of overcrowding problems in hospital emergency rooms or urban transportation).