Indoor Fire Dataset with Distributed Multi-Sensor Nodes (Industrial Hall)

The dataset contains 4 incipient fire scenarios (wood, candles, cable, lunts) along with 8 different nuisance scenarios (smoke from a fog machine, deodorant, ethanol, CO release, exhaust gases, green waste, welding, and abrasive cutting) carried out in a (10 x 22 x 8)m^3 industrial hall without ventilation. Each scenario was repeated 3 times in random with background sequences in between to reduce the influence of prehistory. The dataset consists of 248,502 rows and 18 columns and is structures as a continuous multivariate time series. Each row represents the sensor measurements (CO2, CO, H2, humidity, particulate matter of different sizes, air temperature, VOC and UV) from a unique sensor node position ("Sensor_ID") in the industrial hall at a specific timestamp. The columns correspond to the sensor measurements and include additional labels: a scenario-specific label ("scenario_label"), a binary label ("anomaly_label") distinguishing between "Normal" (background) and "Anomaly" (fire or nuisance scenario), a "fire_label" intersecting the anomalies into fire-relevant or non-fire-relevant anomalies and a progress label ("progress_label") that allows for dividing the events into sub-sequences based on ongoing physical sub-processes. The "Sensor_ID" column can be utilized to access data from different sensor node positions.