Spinsplit’s SpinFlow architecture was used to realize a complex LED (light emitting diode) characterisation system for Budapest University of Technology and Economics. The delivered system boosted the efficiency of LED aging and lifetime modelling studies  within the H2020 AI-TWILIGHT project.

The LED measurement system is a dual-purpose aging and characterization environment designed to autonomously and simultaneously handle up to 20 LEDs. The current implementation uses Spinsplit’s SpinFlow architecture and integrates a thermostat, spectroradiometer, photometer, LED multiplexer, thermal transient tester, and two power supplies into a single cohesive system.
The entire system is managed through a unified user interface called Spinstudio, which not only provides direct control of the hardware components for individual measurements but also facilitates the execution of complex aging workflows.

Spinstudio collects results from all devices in their respective output formats, using a customizable naming convention and timestamps. It organizes these results into folders to facilitate further evaluation and modeling. The collected data includes heating and cooling curves from thermal transient testing, time-dependent radiant flux curves with a high sampling rate throughout the entire thermal transient test to capture the early optical turn-on transient of the device under test. Additionally, spectral power distribution is captured in the steady state, along with various monitoring parameters such as voltages, currents, and ID values of faulty samples to ensure the system’s automated operation. This data serves as the primary input for lifetime-dependent LED modeling used in AI-TWILIGHT and beyond.