© Fotolia, 2012
Manufacturing micro- electromechanical systems, or MEMS as they are known, is a vast and complex process involving sometimes hundreds of different steps, each controlled by a dozen or more parameters, including varying pressures, temperatures and material compounds, etc. Customer needs play a key role in terms of what MEMS are developed, and to what specifications.
The auto industry, for example, uses a range of MEMS like micro-scale accelerometers to trigger airbags or sensors to keep vehicles in line, each combining electrical and mechanical functions with tiny embedded computers etched out of silicon wafers. This 'marriage' of the electro-mechanical and microchip technology is not always a happy one.
Engineering is traditionally wedded to the world of visible, moving parts. But with progress in micro-computing and electronics we are asking MEMS-related product engineering (PE) and electronic design automation (EDA) to get much cosier with developments in micro- and nanotechnology.
This relationship forms something of a chicken and egg problem, suggests Kai Hahn, an expert in the field at Siegen University. Because unlike PE for integrated circuits, the inherent structure, or so-called "third dimension" needed to design MEMS calls for potentially wholesale changes to technology parameters. To resolve this, a deep understanding of the entire PE process for MNT and MEMS is critical. But no one has achieved this
The EU-funded Corona project is the first to develop an integrated design flow, taking into account the process-design stages from product idea to manufacturing and with special emphasis on the end-to-end needs of customers and small MEMS manufacturers in the value chain.
"When Corona started in 2008, there was no dedicated PE methodology for MNT. The tool supply for this high-tech segment was also very poor, so we definitely saw an opportunity," notes Dr Hahn, a key researcher in the consortium. With partners representing key stages along the MNT PE chain, Corona had a head start on competing research groups. It also benefited from an earlier European project, called Promenade, which built software to support the design of MEMS manufacturing sequences.
The team took Promenade's work further by linking its new methodology and tools to current commercial standards, making Corona's MNT PE more user-friendly. "This was important because the customer is the only one who really knows the exact product specifications and can decide on go/no go gates within the PE process," says Dr Hahn.
Corona has achieved all its main objectives: methodology for all design-processing stages of MNT PE; software, middleware and applications supporting the methodology; and real-life MEMS demonstrations.
"The demos carried out by our partners XFAB (Erfurt), ITE (Warsaw), ELMOS (Dortmund), Theon (Athens) and Cambridge University validated Corona's approach and were very helpful to improve our methods and tools," confirms the researcher.
Commercialisation of several tools emanating from Corona is underway. Notably, project partner Coventor (Paris) has commercialised its clever design simulator ('SEMulator3D'). Another partner, Process Relations (Dortmund), has developed 'XperiDesk' for managing myriad design collaboration tasks, from idea to rapid prototyping.
Meanwhile, several prototypes, such as the 'Hedoris' platform developed by academic partner ITE and 'ProcessRecommender' suite by University of Siegen, are undergoing further research. Corona's 'Electronic Product Engineering Flow Manager' is also being put to good use internally by its creator, the firm ELMOS.
IVAM, the project's coordinator, is putting its connections as the industrial association for MNT to good use by communicating Corona's results to its member community. The wider public can also read about the project's achievements and methods in a new book due to be published by Springer in 2012.