NEHIL NEUROMORPHIC-ENHANCED HETEROGENEOUSLY-INTEGRATED FMCW LIDAR

  • I.P.: Miguel C. Soriano
  • Partners: STICHTING IMEC NEDERLAND, INTERUNIVERSITAIR MICRO-ELECTRONICA CENTRUM, THALES, CSIC
  • Fecha de inicio: 1 de Febrero de 2025
  • Fecha de final: 31 de Enero de 2028

The NEHIL project, an EU-Korea partnership, is set to transform the landscape of digital technologies through groundbreaking neuromorphic architectures and advanced heterogeneous integration such as LiDAR systems. This collaborative initiative aims to develop two innovative neuromorphic computing architectures that are crucial for tackling the complex demands of modern data-intensive applications. The first system utilizes FeFET-based Compute-in-Memory (CIM) accelerators, which are designed to support hybrid models of SNN and ANN. These accelerators enhance processing speeds and reduce power consumption, making them ideal for real-time, high-resolution data processing challenges like those found in autonomous vehicle navigation. The second system employs photonic integrated circuits based on reservoir computing (RC) principles, significantly easing manufacturing while enhancing the processing of dynamic data streams. The work involves integrating these neuromorphic systems with state-of-the-art FMCW LiDAR technologies. This integration aims to overcome traditional challenges such as high energy consumption and environmental sensitivity, setting new standards for resolution, accuracy, and cost-efficiency. Specific targets for the NEHIL project include reducing power consumption in object recognition tasks by 50% and achieving a proof-of-concept for ultra-low latency LiDAR signal processing using the FeFET-based CIM and RC architectures. With this approach we exploit the LiDAR's high-resolution capabilities in adverse weather conditions; reducing power consumption, packaging size and manufacturing cost compared to the state of the art. This collaboration extends its benefits beyond the automotive industry, enhancing capabilities in diverse sectors such as telecom, healthcare, smart cities, security, predictive maintenance, infrastructure management, and industrial automation.

Investigadores

  • Miguel C. Soriano

    Miguel C. Soriano

  • Lucas Lacasa

    Lucas Lacasa

Publicaciones recientes

Adaptive control of recurrent neural networks using conceptors

Pourcel, Guillaume; Goldmann, Mirko; Fischer, Ingo; Soriano, Miguel C.
Chaos 34, 103127 (2024)

Esta web utiliza cookies para la recolección de datos con un propósito estadístico. Si continúas navegando, significa que aceptas la instalación de las cookies.


Más información De acuerdo