RT-FDD Datasets: Pick-and-Place and Electric Furnace Benchmarks for Real-Time Fault Detection and Diagnosis

Four benchmark datasets for Real-Time Fault Detection and Diagnosis (RT-FDD) research on discrete manufacturing machines. Each machine (Pick-and-Place and Electric Furnace) is provided in two versions: an Original Dataset (OD) with deterministic fault conditions and an Aleatory Simulated Dataset (ASD) with bounded variability emulating realistic industrial uncertainty. Data was logged from a closed-loop real-time simulation running on a SoftPLC stack (Modbus TCP + OPC UA), organised per operating cycle, preserving event timing, execution order, and continuous sensor behaviour. All four CSVs use Unix epoch milliseconds (UTC) in the timestamp column. See the README for the column reference and the accompanying article for full methodology.