The project “Optimised production of compound semiconductors for increased efficiency in energy supply MOCVD4.2” was carried out as a consortium effort between AZUR SPACE Solar Power GmbH, AIXTRON, Laytec, RWTH Aachen and the Forschungszentrum Jülich. Funded by the Projectträger Jülich (PTJ) the work spanned the period from early 2022 to mid‑2023, with the final report dated 23 May 2023 and approved on 28 June 2023. AZUR provided the industrial context and use‑case definition, AIXTRON supplied the MOCVD hardware, Laytec contributed sensor and software modules, RWTH Aachen offered system‑integration expertise, and the Forschungszentrum Jülich coordinated the overall project and managed the funding.
Technically the consortium developed a next‑generation MOCVD 4.0 process platform that embeds Industry 4.0 principles. The first phase focused on defining the required technology and sensor suite, drawing on AZUR’s production experience to ensure that the new platform would meet the semiconductor industry’s demands for reproducibility and flexibility. From these use cases a comprehensive system specification was drafted, detailing a modular architecture that integrates reactor control, sensor data acquisition, and edge‑computing analytics. The specification was reviewed and frozen jointly by AZUR and AIXTRON, marking the successful completion of work packages 2.1 and 2.2.
Work package 2.4 addressed the networking layer and security concept. AZUR’s team guided the partners in designing interface specifications that allow secure, real‑time communication between the reactor, sensors, and the central control software. The resulting architecture supports modular plug‑in of new sensors and firmware updates without disrupting ongoing production.
The core of the project was the system integration and testing phase (work package 2.8). Here the consortium evaluated the functionality and usability of the hardware and software developed by AIXTRON and Laytec. A key outcome was the implementation of a post‑maintenance diagnostic routine: after each maintenance cycle the software automatically runs a predefined set of reactor conditions, records the results, and compares them to the last stable state. This routine provides a quantitative baseline that enables early detection of deviations, thereby reducing failure rates and improving yield. While the report does not provide explicit numerical performance figures, it emphasises that the new diagnostic capability markedly enhances process stability and reproducibility.
Overall, the project demonstrated that integrating Industry 4.0 technologies into MOCVD processes is feasible and yields tangible benefits for semiconductor manufacturing. The new MOCVD 4.0 platform offers a flexible, modular framework that can accommodate a wide range of material systems and product types, addressing the need for rapid adaptation to changing market demands in power electronics, photovoltaics, and optoelectronics. The collaboration model—combining industrial know‑how, equipment manufacturing, sensor integration, and academic system engineering—proved effective in translating conceptual requirements into a working production system. The consortium’s results lay the groundwork for further optimisation of compound‑semiconductor production and provide a blueprint for future Industry 4.0 implementations in the semiconductor sector.