Do today’s neural networks have the imagination necessary to address the requirements and constraints of high-tech manufacturing roles?
As fabs struggle to keep up with the challenges associated with quality controls found at the intersection of Moore’s Law and microelectronics manufacturing, automation and digitization strategies are increasing.
Dr. Gerhard Luhn and André Schaaf, together with Infineon 300 mm Ramp engineer Dr. Germar Schneider, presented recent research findings from neuroscience on imagination, to challenge and facilitate the role of humans and human knowledge in future fabs at the APC Conference XXX 2018.
Integrated Development 4.0 (iDev40)
The 47 million Euro iDev40 research project is focused on networking for development and production processes for electronic components and systems.
Within this project, the connection to SYSTEMA’s Universal Signature Engine (USE) is clear given the challenges of fast and qualitatively reliable development and release for production of newly developed technologies while discovering new types and categories of failures. Read more
USE’s features perfectly address those challenges:
During everyday work, new failures occur which require the implementation of new algorithms and methodologies. USE provides a framework which efficiently accommodates the integration of such algorithms and classification methodologies.
USE also provides evaluation and optimization modules and ensures high quality standards for the created models. With a user-friendly interface as well as visual script creation, USE makes it easy for the operator/engineer to create new scripts, maintain, optimize, test and debug. The production backend then provides automatic classification and defect code assignment in production lines 24/7.
The APC Conference allowed us to demonstrate that the benchmarks to setup new productive requirements can be successfully met within a single working day. Even adapting new algorithms and methodologies like neural networks and deep learning are covered within the consolidated approach.
We had many interesting conversations with industry professionals about neural networks, deep learning and quantum computing. This exchange brought us new insights into the current state of deep learning technologies, which led to new requirements and fresh ideas to move forward in the future. Read more
38 project partners from six countries cooperate in iDev40.