Establishing TS-FM Baseline Standard

Background: The transition to Time Series Foundation Models (TS-FMs) marks a new operational paradigm in time series forecasting, contrasting with traditional dataset-specific deep learning approaches.

Question / Future Work: Future research in transportation forecasting should establish Time Series Foundation Models, such as Chronos-2, as default, competitive baselines against which newly proposed, domain-specific architectures must be benchmarked to justify their added complexity.