NeAIxt

Project overview

NeAIxt is a three-year European research and innovation project developing the next generation of secure, energy-efficient, high-performance Edge AI, making an important contribution to Europe’s digital and industrial future.

The project is coordinated by STMicroelectronics and co-funded by the CHIPS Joint Undertaking under Horizon Europe.

With 66 partners from 14 European countries—including large enterprises, SMEs, research and technology organisations, and academic institutions—NeAIxt brings together extensive expertise across the entire microelectronics and artificial intelligence value chain.

 

Project objectives

The aim of NeAIxt is to scale and further develop key technologies for artificial intelligence at the network edge (Edge AI):

• Scaling up: Broad industrial applicability and high market readiness
• Enhancing: Improved accuracy, performance, and energy efficiency
• Demonstration: Deployment of the developed technologies in large volumes and real industrial applications

In doing so, NeAIxt strengthens Europe’s technological sovereignty and enables reliable AI functionality directly on the end device (“on-device”).

 

Technological focus areas

NeAIxt addresses key hardware and software enablers for Edge AI systems:

• Embedded Non-Volatile Memory (eNVM): Further development of the eNVM roadmap with a focus on advanced technologies such as ePCM, OxRAM, and FeRAM, optimised for AI workloads
• FD-SOI-based chip architectures: Foundation for particularly energy-efficient, secure, and high-performance Edge AI solutions
• Edge AI hardware and software: Development of specialised microcontroller architectures (MCUs), including a coordinated software ecosystem of frameworks, tools, and libraries
• Low-Power & Security: Architectures for extremely low power consumption and secure AI processing directly at the network edge

 

microsensys’ contribution

Battery-powered wireless sensors play a central role in applications where high flexibility and strong miniaturisation are required. In particular, in the transport sector they capture and monitor a wide range of physical parameters. Since wireless transmission causes the greatest energy demand compared to measurement data acquisition, it is crucial to activate the wireless functionality only when it is actually needed—for example, when a suitable receiver is within range.

As part of the NeAIxt research project, microsensys GmbH is therefore developing, together with project partners, a novel sensor that can be selectively woken using passive UHF RFID technology and transmits the captured measurement data via a battery-powered BLE SoC. This is based on a specially developed RFID ASIC that uses particularly fast and energy-efficient FeRAM memory technologies. This enables the sensor to be used even on fast-moving objects, where only short time windows are available for activation and data transmission.

To further minimise power consumption and ensure particularly fast data transmission, embedded ML methods for data reduction are also used directly in the sensor. This results in a highly integrated Edge AI system that provides the basis for long-lasting, energy-efficient, and responsive sensor networks.

If you would like to learn more about our project, please visit our project website.