AI-DISCO

RelationshipCollaborative ProjectStatusActiveFunderFederal Ministry of Research, Technology and Space (BMFTR)Duration2026-2030
AI-DISCO project image

Project overview

* motivation Research-intensive microelectronics and its applications are drivers of progress, competition, and innovation across all industries. To accelerate the transfer of knowledge and expertise to industry and to support structural change in the Lusatian lignite mining region, existing research structures and competencies in microelectronics in Cottbus will be significantly expanded. * Objectives and approach The aim of this project is to develop a new generation of sensors that not only collect data but also process it intelligently. The project partners are miniaturizing hardware for artificial intelligence in an innovative circuit architecture to integrate it directly into the sensors. They are developing optimized circuit designs that consume less energy and react faster, thereby increasing the efficiency of devices and machines. New approaches in the field of neuromorphic computing are also being pursued. The developed circuits and demonstrators will be tested in environmental monitoring applications and smart buildings in collaboration with industry partners to ensure their practical applicability. The goal is to create a foundation for autonomously operating systems that can process data efficiently and react quickly to changes. * Innovations and perspectives The drastic miniaturization and increased energy efficiency of AI hardware allow the typically external processing of sensor data to be moved directly into the sensor itself. This makes the sensors intelligent and reduces the amount of data that would otherwise be transmitted to a cloud computing infrastructure and the user. The planned hardware platform is designed to be flexible in order to support as many sensor types and applications as possible.

active
Funder
Federal Ministry of Research, Technology and Space (BMFTR)
Duration
2026-2030
Partners
* Brandenburg University of Technology Cottbus-Senftenberg * German Research Center for Artificial Intelligence GmbH, Kaiserslautern * Fraunhofer Institute for Photonic Microsystems IPMS, Dresden * Leibniz Institute for Neurobiology (LIN), Magdeburg * Perinet GmbH, Cottbus * Enertrag SE, Schenkenburg * ABB, Cottbus site
Grant ID
16ME1127
Budget
€15.6 Million

Institutional context

  • Leibniz Institute for High Performance Microelectronics

    Consortium coordinator

  • Brandenburg University of Technology Cottbus-Senftenberg

    Consortium institution

  • German Research Center for Artificial Intelligence

    Consortium institution

  • Fraunhofer Institute for Photonic Microsystems IPMS

    Consortium institution

  • Leibniz Institute for Neurobiology

    Consortium institution

  • Perinet GmbH

    Consortium institution

  • ENERTRAG SE

    Consortium institution

  • ABB AG

    Consortium institution

Publications and outputs

  1. 20260 citations

    REQAP: Resilient Weight Packing and Quantization for Edge DNN Acceleration

    Mahdi Taheri, Samira Nazari, Mubassher Ansari, Ali Azarpeyvand, Mohsen Afsharchi, Maksim Jenihhin, and Christian Herglotz