The European microelectronics industry has a strong footprint in Automotive, Medical and Industrial markets, which are expected to grow significantly within the coming years. Driven by the transition towards connected, autonomous, shared and electric mobility, by the trend to personal medical devices and the needs of an aging society, as well as by the digital transformation of manufacturing, the requirements for microelectronic products are increasing, which calls for innovative solutions for IC manufacturing and design enablement and verification. The presentation is touching on process and device co-design challenges by using wafer scale CMOS and MEMS technologies in the development of system solutions. Using examples from automotive, medical and industrial, it will be shown how novel design technology can help to de-risk the design process, to reduce re-designs and time-to-market and to enable new applications.
Jörg Doblaski is Chief Technology Officer at X-FAB, and in this role responsible for the foundry’s process technology development, design enablement and technical support. Prior to the CTO role, Jörg served in different engineering- and engineering management positions, with focus on the area of design support and design enablement. Jörg studied at the Technical University of Ilmenau, and holds a diploma degree in Electrical engineering and information technology.
The design and characterization of microelectronic systems are substantially driven by experience and knowledge of design and test engineers. For the most part, such knowledge can neither be formally mapped nor used by methods of automation. Therefore, suboptimal solutions in complex analog or mixed analog / digital systems are often only recognized late. This causes additional effort ultimately resulting in high costs in design and validation of microelectronic systems after production. Meanwhile, machine learning algorithms achieved outstanding capabilities in areas such as natural language and image processing. Integrating learning algorithms into the design and characterization process of microelectronic structures, such techniques offer the potential for anomaly detection, data compression, and design optimization. The talk provides an overview on different application scenarios of using machine learning techniques for design and characterization of microelectronic systems. Learning models for the creation of behavioral predictions of system components, the automatic analysis of characterization data, and for anomaly detection in layout designs are discussed.
After receiving the M.Sc. degree in Technical Physics from the Technische Universität Ilmenau in 2012, and the Ph.D. degree in Physics in 2015, Marco Seeland fulfilled his long-standing passion and started as a postdoc researcher in the field of computer science. After joining the Software Engineering for Safety-critical Systems group at the Department of Computer Science and Automation, Technische Universität Ilmenau, he has been working in and leading several projects with a focus on data science and machine learning in different domains. His main research interest is development and application of neural networks for computer vision and pattern recognition in ecological bioinformatics to enable citizen science applications such as Flora Incognita.