- 03/11/2026
- Expert knowledge
- Event
Winners of the embedded award 2026
A high-calibre jury evaluated more than 110 entries for the 22nd embedded award. We present the winners in the nine categories of the embedded award and the additional category of the Community Choice Award – which was chosen by the embedded community itself.
Artificial Intelligence
Ceva-NeuPro-Nano
Exhibitor: Ceva
Hall/Booth: 4-462
Ceva NeuPro Nano is a fully self sufficient Neural Processing Unit (NPU) that brings powerful, energy efficient AI directly to highly resource constrained edge devices. While traditional solutions depend on additional host CPUs or DSPs—adding complexity, power consumption, and cost—NeuPro Nano integrates neural inference, control logic, signal processing, and memory management into a single programmable NPU core.
Supporting modern ML data formats from 4 to 32 bit, native transformer processing, hardware accelerated sparsity, and fast quantization, NeuPro Nano delivers scalable performance with exceptionally low energy use. The proprietary Ceva NetSqueeze technology enables direct processing of compressed weights, reducing model memory requirements by up to 80 %.
With ultra low power operation, deterministic latency, and a compact system architecture, NeuPro Nano is ideal for always on applications in consumer IoT devices, industrial sensing, and automotive systems. Its efficiency extends battery life, reduces material needs, and minimizes cloud dependency through local AI inference.
NeuPro Nano offers a future proof, sustainable, and cost efficient platform for physical AI applications right at the sensor—delivering maximum performance with minimal resources.
Electronic Displays
Adaptive E-Paper Driving
Exhibitor: LCD Mikroelektronik GmbH
Hall/Booth: 1-351
Adaptive E-Paper Driving redefines how E-Paper displays can be controlled, using purely software to achieve high-quality grayscale rendering, faster updates, and optimal energy efficiency. Unlike conventional solutions that rely on static, manufacturer-defined waveforms, this technology generates dynamic, adaptive driving waveforms (LUTs) in real time—bringing even simple 1- or 2-bit controllers to performance levels previously possible only with additional hardware.
The solution shows on a 4.2-inch prototype display, which would normally render only black and white, that up to 16 grayscale levels can be displayed with noticeably improved detail. By removing the need for extra hardware, the software-driven approach reduces system complexity, power consumption, and costs, while allowing precise adaptation to customer requirements such as grayscale depth, update speed, energy profile, or operating temperature. Although still in a prototype phase, Adaptive E-Paper Driving shows the potential to unlock new possibilities in digital signage, industrial visualization, and sustainable information systems—wherever highly flexible, energy-efficient, and long-lasting displays are needed.
Embedded Vision
Memorence Operagents
Exhibitor: Memorence AI
Hall/Booth: 3-439
Memorence Operagents brings real-time intelligence directly to frontline operations. The system assists industrial and service personnel by understanding visual context, user intent, and procedural steps in real time. This ensures reliable execution of complex SOPs, reduces human errors, and adapts instantly to changing conditions without requiring retraining.
At its core is an instant-learning on-device AI developed by Memorence AI for real-time operational environments. Operagents integrates embedded vision, multimodal sensing, and adaptive learning within a compact embedded platform. The platform also incorporates vision-language models (VLM) and natural voice interaction, enabling contextual understanding and explainable guidance directly on site. It delivers step-by-step guidance, immediate validation, and contextual explanations—fully offline, low-latency, and compliant with strict data-privacy requirements.
Compared to traditional AI solutions that stop at detection or analytics, Operagents closes the loop between perception and action within operational workflows. Operator corrections are applied immediately, enabling rapid adaptation alongside conventional retraining processes.
The system can be deployed in real production environments without requiring dedicated fixtures or controlled lighting enclosures. Memorence Operagents increases quality and consistency by preventing errors early in the workflow and supporting reliable execution. The result is a practical and scalable human–AI collaboration framework for reliable industrial operations.
Hardware
eiscube & ioneis cloud
Exhibitor: ioncentric
Hall/Booth: 5-282
The eiscube closes a critical gap in today’s battery and electronic systems by enabling Electrochemical Impedance Spectroscopy (EIS) directly in the field. While traditional methods such as voltage or DCIR measurements provide only rough status indicators, EIS “looks” inside the cell and detects degradation, reaction resistance, and safety relevant changes at an early stage. This transforms maintenance from reactive to truly predictive and significantly extends the lifetime of electronic and electrochemical systems.
All measurement data is processed directly on the device, cryptographically signed, and transmitted in a tamper proof manner. Through the ioneis cloud, the data feeds into analytics, digital product passports, and automated lifecycle documentation.
The combination of EIS intelligence at the edge and zero trust communication is unique in the market. It enables precise, autonomous diagnostics outside of laboratory environments—for batteries, converters, inverters, and motors. A European patent is currently being prepared to protect the underlying process.
With modular, repairable hardware and trustworthy lifecycle data, the eiscube supports sustainable value chains, reduces material consumption, and extends the usable life of industrial assets.
Safety & Security
Exein Analyzer
Exhibitor: Exein
Hall/Booth: 5-253
Exein Analyzer transforms how manufacturers prepare for security and product certifications such as the EU Cyber Resilience Act. Until now, teams had to manually gather documents, map evidence to requirements, and identify gaps. Analyzer automates this entire workflow. The system analyzes technical artefacts—including firmware, SBOMs, vulnerabilities, and configuration data—alongside organizational documentation such as architecture descriptions, test reports, and security procedures. All inputs are automatically classified, matched to regulatory requirements, and checked for completeness.
This creates a centralized, audit ready evidence inventory that clearly shows which requirements are fulfilled, where evidence is missing, and how gaps can be resolved. Analyzer is the first solution to combine deep firmware analysis with documentation based compliance assessment in a single workflow—unlike tools that address only isolated aspects or rely on expensive manual consulting.
Output includes complete compliance reports, coverage summaries, gap analyses with prioritization, and actionable recommendations. Exein Analyzer reduces certification timelines from months to weeks while significantly lowering cost and risk. Integrated into CI/CD pipelines, it enables continuous compliance monitoring and early detection of security issues—supporting more sustainable development processes and longer lasting, more secure products.
SoC / IP / IC Design
ZF I/O Interface Chip for Automotive HPC with SiliconAuto Companion Controller World’s First Live Demonstration of Real-Time Sensor Data Acquisition and PreProcessing on Silicon to Enable AD up to L4
Exhibitor: ZF Friedrichshafen AG powered by SiliconAuto
Hall/Booth: 4-665
The new ZF I/O Interface Chip, paired with SiliconAuto’s Companion Controller, transforms the architecture of future automotive high performance computers. Instead of concentrating all sensor interfaces and preprocessing tasks on expensive performance SoCs—often leading to CPU overload and memory bottlenecks—the chip performs real time acquisition and preprocessing of camera and radar data directly on silicon. This frees valuable CPU resources, avoids memory contention, and ensures reliable scaling of multiple parallel sensor interfaces.
Its core innovation is a highly integrated I/O platform that incorporates all essential automotive interfaces (CSI 2, radar signal processing, Ethernet, and more) with deterministic streaming based processing. Coupled with SiliconAuto’s XMotiv M3 for fast boot, power sequencing, system supervision, and secure boot, the solution delivers immediate readiness and robust control. Through standardized high speed links such as PCIe or UCIe, the architecture remains fully SoC agnostic, giving OEMs maximum flexibility in selecting their compute platform.
The USP: the world’s first live demonstration of a specialized I/O chip with real time camera ISP and radar preprocessing integrated directly in hardware. Precise timestamping, deterministic synchronization, and an optimized DRAM/SRAM architecture reduce latency, cost, and power consumption—while ensuring “turn key” sensor availability right after ignition.
As a modular and scalable concept, the solution supports future chiplet based HPC designs, open standards, and sustainable upgrade paths without costly redesigns—paving the way for energy efficient, sovereign automotive computing in Europe.
Software
eXtremeDB/rt
Exhibitor: McObject
Hall/Booth: 4-238
eXtremeDB/rt is the world’s only database that fully meets hard real time requirements. Traditional database systems, flash file systems, and flash translation layers lack any concept of deadlines and therefore cannot guarantee deterministic behavior. eXtremeDB/rt solves this by integrating a real-time–capable transactional flash translation layer directly into the database kernel, creating a fully deadline aware software stack from the application down to the NAND flash hardware.
The system introduces several technological breakthroughs: transactions include explicit deadlines; a real time scheduler aligns with EDF or priority based RTOS policies; transaction progress is continuously monitored to ensure deterministic aborts; and a copy on write logging mechanism replaces nondeterministic write ahead logging. As a result, even large and complex datasets can be processed reliably—especially in systems where RAM is insufficient and flash memory is unavoidable.
Its unique advantage: eXtremeDB/rt is the only commercially available hard real time DBMS. Other so called “real time databases” merely analyze streaming data and cannot support actual real time constraints.
eXtremeDB/rt also contributes to sustainability: by eliminating the flash file system and reducing write amplification, flash wear is minimized, extending device lifetime and reducing electronic waste.
Startup
M8 operating system
Exhibitor: Sqrrl
Hall/Booth: 4-604
The M8 operating system bridges the growing gap between traditional real-time operating systems and high-performance general-purpose OS. It combines deterministic real-time behaviour with the flexibility of modern Linux-based development—delivering a single, certifiable platform for automotive, aerospace, industrial, and defence applications.
Its innovative microkernel design guarantees the strictest isolation for the highest safety and security requirements. But it also supports native POSIX and Linux APIs to target multi-core heterogeneous systems-on-chip, enabling frictionless reuse of Linux code from legacy and/or prototype solutions without time consuming rewrites.
Virtualisation and containerisation are also natively supported. But they are designed as integrated into the strict M8 partitioning logic so that containers enjoy full isolation exactly as virtual machines. This unique design guarantees a uniquely flexible approach to the freedom from interference required for ASIL D, SIL 4, or DAL A certifications.
The result is a new safe and secure operating system. M8 is a platform offering 10× higher computational efficiency and up to 50% lower integration costs. With fully European IP sovereignty, modular architecture, and native virtualisation, M8 enables the fastest consolidation of multiple ECUs onto a single SoC—resource efficient, energy optimised, and built for long-term sustainability.
Tools
Ambiq HeliaAOT
Exhibitor: Ambiq Micro
Hall/Booth: 4-581
Ambiq HeliaAOT is an ahead of time compiler solution that enables modern AI models to run efficiently on highly resource constrained edge devices. While traditional AI runtimes introduce unnecessary overhead, high memory demands, and unpredictable memory usage, HeliaAOT converts trained models—such as TensorFlow Lite—into highly optimized, standalone C inference modules already at build time.
By eliminating a generic runtime, removing unused operators, and providing automatic, deterministic memory planning, HeliaAOT produces extremely compact, transparent, and energy efficient deployments. Hardware optimized transformations such as layer fusion, tensor reordering, and graph based optimizations significantly reduce code size and memory footprint while still enabling powerful, real time inference performance.
Developed specifically for Ambiq’s ultra low power SoCs with SPOT® technology, HeliaAOT delivers up to 10× smaller binaries, predictable execution, and industry leading efficiency—ideal for wearables, hearables, sensors, and other always on devices. With reduced energy consumption, smaller memory requirements, and extended device lifecycles, HeliaAOT also contributes to sustainable, resource efficient edge AI systems.