Transient Electro-Thermal Measurement Dataset of OSRAM Oslon Black Flat Automotive LEDs

This dataset contains synchronized electro–thermal measurement data of an automotive LED module equipped with OSRAM Oslon Black Flat LEDs. The measurements were conducted to capture the transient electrical and thermal behavior of the device under dynamic excitation. The dataset is designed for research on data-driven electro–thermal modeling and time-series forecasting of LED systems. It consists of six independent measurement runs, each representing a complete transient sweep performed under different experimental settings. During each run, the LED module was excited using a pseudo-random multilevel current profile that drives the device through a wide range of operating conditions. This excitation strategy allows the electro-thermal system to naturally traverse a broad region of the current-voltage-temperature space within a few experiments. For each run, the measurement data is stored in two separate files generated by different measurement devices:

- Electrical measurement files (Evaluation_*.json) contain the electrical measurements acquired by the source-measure unit. Each file includes both measurement metadata and time-series arrays. The recorded signals include: voltage (terminal voltage of the LED module), current (applied drive current), time stamps, sample indices, and some metadata describing the measurement configuration, such as excitation parameters, compliance voltage, measurement duration, and other acquisition settings.
- Thermal recording files (recording_*.csv) files contain the thermal measurement data, recorded independently using a thermal imaging system. The recordings include temperature measurements of the LED chip region (R2) and the surrounding environment (R1) obtained from infrared thermography. The thermal data is stored as time-series values together with corresponding time stamps.

Because the electrical and thermal measurements originate from two independent acquisition systems, their time bases are not synchronized. As a result, the timestamps of the electrical (Evaluation_*.json) and thermal (recording_*.csv) files differ slightly in both start time and sampling rate. To combine both signals for analysis or modeling, the data must therefore be temporally aligned and resampled based on detecting the current trigger and temperature change.