PRIME was a Marie Sklodowska-Curie Individual Fellowship project, (H2020-MSCA-IF-EF-ST, GA 101032925) with the vision to introduce a fresh perspective on the assessment of high-power RF MEMS reliability. The way to achieve this was by combining conventional reliability testing with machine learning techniques towards enabling failure related predictive diagnostics. Nevertheless, the project was a holistic approach and apart from the machine learning reinforced studies, also included RF-design tasks, micro/nano-fabrication activities, and implementation of diverse characterization techniques having always in mind to support the development of the fellow’s profile, to maximize the transfer of knowledge between the fellow and the host and to disseminate the results to the broader possible audience.

Towards this end, PRIME workplan included the design of the devices to be studied, their fabrication, extensive characterization, monitoring of the RF response and power handling capabilities, study of aspects related to reliability physics and development of machine learning based methodologies enabling predictions concerning their performance and reliability. All these tasks were performed in the facilities of the host (IESL-FORTH) but also through two secondment placements related to RF design (RF-Microtech, Italy) and to machine learning tasks (Imperial College London, UK).

Through this workplan PRIME achieved to deliver a family of RF MEMS switches supporting different RF responses and power handling capabilities. In addition, reliability physics issues anticipated to appear during high power operation of RF MEMS were identified and studied. Finally, machine learning based methodologies utilising datasets generated by either simulation studies or by experimental results have been introduced.

Beyond the purely research tasks, PRIME was also focusing on the fellow. Therefore, the workplan was designed in such a way as to offer the complementary skills that were missing from the fellow’s profile. This included research skills, managerial tasks as well as the establishment of international network of collaborators, fact highlighted by the undertaking by the fellow of the coordination on behalf of FORTH of a new KDT JU project. In parallel through PRIME workplan the fellow also transferred to members of the host useful knowledge about RF MEMS physics, reliability, and characterization.

PRIME research outcomes so far resulted in the following publications.

Peer Reviewed Journal

• L. Michalas et al. “RF signals over field emission currents: A theoretical study for MEMS capacitive switches” Microelectronics Reliability 138, 114678, 2022.

Conference Proceedings

•  L. Michalas et al. “RF MEMS switch design methodology by electromagnetic simulations and machine learning” 2022 52^nd European Microwave Conference (EuMC), Milan, Italy, 2022, pp. 369-372. 

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No [101032925].