Consumer
Consumer products increasingly rely on intelligent, responsive and energy‑efficient technologies to deliver natural user interaction, real‑time processing and enhanced privacy. By bringing AI directly to devices such as audio interfaces, connected objects and sensing systems, NeAIxt enables low‑latency, privacy‑preserving consumer experiences that operate reliably under strict power and resource constraints.
Through 5 Consumer‑focused Ecosystems, NeAIxt translates these needs into concrete Edge‑AI solutions, enabling ultra‑low‑power audio processing, intelligent wireless communication and sensor‑based interaction for next‑generation consumer devices.
Ecosystem : Ultra-low power speech enhancement
This Ecosystem develops highly‑compressed neural speech‑enhancement models capable of real‑time noise cancellation on ultra‑low‑power edge processors, enabling next‑generation hearing and audio devices.
| Ecosystem Partner list | Activity details | |
|
|
GN Hearing |
Ecosystem and Use case leader: GN provides advanced speech‑enhancement models, audio‑denoising algorithms, and leads the deployment/optimization pipeline to run these models efficiently on constrained edge devices. |
|
|
University of Thessaloniki |
AUTH develops deep‑neural‑network compression, extensions, optimized kernels, and customizations enabling compressed speech‑enhancement models to run on ultra‑low‑power AI accelerators. |
 |
Technical University of Denmark |
DTU contributes embedded‑AI frameworks and energy‑efficient design methods to support real‑time speech enhancement on resource‑constrained hardware platforms.
|
Ecosystem : AI-managed RF digitizer for Cognitive Radio, Software Radio, 6G, AIoT
This Ecosystem develops a new generation of AI‑enabled wireless technology that can automatically identify and use the best available radio frequencies in real time, while converting radio signals directly inside the device.
| Ecosystem Partner list | Activity details | |
|
|
Instituto de Microelectronica de Sevilla |
Ecosystem and Use case leader: IMSE-USE designs the core wireless signal converter and embeds AI directly into it, enabling the device to understand its radio environment, learn from data and adapt its behavior in real time. |
|
|
STMicroelectronics CROLLES |
ST CRO provides the underlying chip technologies that make AI‑enabled wireless communication possible, and supports their integration into compact, energy‑efficient devices. |
 |
STMicroelectronics GRENOBLE | STE GRE provides an AI‑ready microcontroller platform that serves as the foundation for testing and running intelligent features directly on consumer devices. |
 |
TUBITAK | TUBITAK brings AI models onto embedded consumer platforms, ensuring they can run efficiently on low‑power hardware. This includes adapting AI models for embedded use, supporting the development tools, and enabling intelligent analysis of wireless signals directly on the device. |
Ecosystem : AI-Driven Intelligent Edge Wireless Communication
This Ecosystem develops AI‑enabled cognitive‑radio and signal‑processing blocks for ultra‑low‑power edge devices, enabling 6G‑class wireless systems that autonomously adapt transmission parameters to the radio environment.
| Ecosystem Partner list | Activity details | |
|
TUBITAK | Ecosystem and Use case leader : TUBITAK develops AI‑enabled communication solutions that help devices adapt their wireless behavior in real time while remaining energy‑efficient and suitable for embedded consumer platforms. |
 |
CEA | CEA brings next‑generation memory technologies and system‑level expertise to enable efficient, low‑power Edge‑AI hardware for wireless consumer devices. |
Ecosystem : Sensor fusion for human activity detection in buildings
This Ecosystem enables privacy‑preserving human‑activity detection in buildings by fusing acoustic, Bluetooth Low Energy (BLE) and auxiliary sensors at the Edge, improving occupancy accuracy for smart‑building automation.
| Ecosystem Partner list | Activity details | |
 |
Sorama | Ecosystem and Use case leader : Sorama provides acoustic cameras and processes sound‑field data to enrich occupancy detection and enhance localization accuracy in multi‑sensor setups. |
 |
Almende | Almende contributes IoT nodes with BLE scanning, integrates sensors, and develops edge‑level fusion algorithms for real‑time human‑activity tracking. |
 |
Innatera | Innatera provides neuromorphic processors and spiking‑neural‑network tooling suitable for ultra‑low‑power event‑driven sensing pipelines in multi‑sensor environments. |
 |
IMEC NL | IMEC NL contributes scalable Edge‑AI hardware and co‑optimization methods, enabling efficient mapping of multi‑sensor fusion workloads. |
Ecosystem : Processing of Acoustic Signals by Adaptive Filters and AI Inference
This Ecosystem delivers an ultra‑low‑cost, ultra‑low‑power voice‑control engine running on STM32, using adaptive filtering and AI inference to recognize 20 Czech spoken commands with near‑perfect reliability.
| Ecosystem Partner list | Activity details | |
 |
Institute of Information Theory and Automation |
Ecosystem and Use case leader : UTIA develops an efficient on‑device voice‑control pipeline: noise reduction, AI training for 20 Czech commands, and real‑time command recognition on an embedded microcontroller. |
|
STMicroelectronics CROLLES | ST CRO delivers the chip and memory foundations needed to run AI‑based audio processing efficiently on low‑power consumer devices. |
|
STMicroelectronics GRENOBLE | ST GRE delivers an AI‑enabled microcontroller and development hardware that support efficient, real‑time voice control. |