Dataset for Unsupervised Anomaly Detection in High-Dimensional Vehicle Telemetry using Isolation Forests on Edge Devices

This dataset contains multivariate time-series telemetry data designed to support the development of lightweight predictive maintenance frameworks for small-to-medium fleet operators. The data simulates high-frequency sensor streams from a heavy-duty commercial vehicle’s Controller Area Network (CAN) bus, which is the current industry standard for vehicle diagnostics.