MOSAIC

Magnetic and Cognitive Systems for Resilient Automated Perception

MOSAIC is a large-scale Horizon Europe project addressing a central challenge for European competitiveness: technological sovereignty in automated systems and Electronic Components and Systems (ECS). Closely aligned with the objectives of the EU Chips Act, MOSAIC aims to strengthen Europe’s digital autonomy by advancing next-generation ECS technologies and accelerating their adoption across key industrial sectors.

The project focuses on the development of cognitive, energy-efficient, and robust perception systems tailored for automated platforms. These systems are designed to process large volumes of data in real time, support AI-enabled decision making, and operate reliably in complex and demanding environments. A key ambition of MOSAIC is to enable non-invasive, distributed perception architectures that avoid single points of failure while reducing system complexity and improving accuracy.

To achieve this, MOSAIC addresses challenges spanning hardware, communication, and system integration. The project emphasizes interoperability and standardized communication protocols, fostering a cohesive ecosystem across industries including automotive, aerospace, maritime, industrial automation, and critical infrastructures. Its outcomes will be demonstrated through 31 advanced technical showcases, such as 360° distributed radar perception, AI-enabled reasoning based on magnetic-field signatures, and resilient communications using non-terrestrial networks. Two complementary impact studies will further consolidate Europe’s leadership in automated systems.

MOSAIC brings together a pan-European consortium covering the entire ECS value chain, ensuring that innovations translate into manufacturable technologies and contribute directly to filling European semiconductor fabs.

Our Contribution: Robust Magnetic Sensing for Automated Systems

Within MOSAIC, our group is responsible for the design and development of planar Hall effect (PHE) magnetic sensors tailored for next-generation automated and distributed perception systems.

Our work focuses on creating sensor architectures capable of detecting weak magnetic fingerprints with high sensitivity, while maintaining resilience under harsh operating conditions. These conditions include temperature gradients, mechanical stress and pressure, and exposure to radiation—environments commonly encountered in aerospace, industrial, and infrastructure applications.

To enhance system-level performance, the PHE sensors are co-designed with:

• Dedicated CMOS interface electronics, optimized for low noise and efficient signal acquisition

• Neuromorphic and AI-based algorithms that fuse data from multiple sensors to improve robustness, reliability, and contextual interpretation

This combined approach—integrating sensors, electronics, and algorithms—enables distributed magnetic-field monitoring with improved fault tolerance and reduced susceptibility to environmental perturbations. The resulting solutions are intended to be scalable and adaptable across a wide range of applications, including IoT systems, transportation platforms, and real-time monitoring of industrial equipment.