Sharing, collaborating, and benchmarking to advance travel demand research: A demonstration of short-term ridership prediction

This research foregrounds general practices in travel demand research, emphasizing the need to change our ways. A critical barrier preventing travel demand literature from effectively informing policy is the volume of publications without clear, consolidated benchmarks, making it difficult for researchers and policymakers to gather insights and use models to guide decision-making. By emphasizing reproducibility and open collaboration, we aim to enhance the reliability and policy relevance of travel demand research. We demonstrate this approach in the field of short-term ridership prediction. Drawing insights from over 300 studies, we develop an open-source codebase implementing five common models and propose a standardized benchmark dataset from Bogotá’s transit system, which we use to evaluate these models across stable and disruptive conditions. Our evaluation shows that online training significantly improves the prediction accuracy under demand fluctuations, with the multi-output, online-training LSTM model performing best across stable and disrupted conditions. However, even this model required approximately 1.5 months for error stabilization during the COVID-19 pandemic. The aim of this open-source codebase is to lower the barrier for other researchers to replicate models and build upon findings. We encourage researchers to test their modeling approaches on this benchmarking platform using the proposed dataset or their own, challenge our analyses, and develop model specifications that can outperform those evaluated here. Further, collaborative research approaches must be expanded across travel demand modeling if we wish to impact policy and planning.