The coveted embedded awards were presented for the 18th time on June 21 at 12:00 noon at Embedded World. The jury had selected 16 of the 55 applications received in advance and nominated them for the award.

We present the winners in the six categories on the following pages:

• Embedded Vision
• Hardware
• Safety & Security
• Software
• Tools
• Startup

The winners in the 'Embedded Vision' category

Beaming winners in the Embedded Vision category: Christian Verbrugge (left) and Paschalis Pelitaris (right), both GrAI Matter Labs, are delighted to have won the embedded award 2022. The trophy and certificate were presented to them by Yasmine Blair (right), moderator of NürnbergMesse, and Prof. Dr.-Ing. Axel Sikora (left), Head of Jury.

© Uwe Niklas

GrAI Matter Labs with the Life-Ready AI Chip

GrAI Matter Labs (GML) has developed a new type of data flow computing architecture that combines the principles of biological systems, near-storage computing power and massively parallel cores. They call it NeuronFlow. The calculations within the NeuronFlow architecture are event-based, whereby a neuron operation is only executed when new data arrives at this neuron. New data or events only arrive at the neuron if there have been significant changes compared to the previous point in time.

GML's approach to the NeuronFlow technique makes it possible to avoid unnecessary calculations and only execute the most important aspects of a deep neural network. GML's unique neuromorphic computing paradigm provides massively parallel and fully programmable sensor analytics and machine learning with significantly reduced power consumption. GML's programmable application processors provide AI with machine learning for the Internet of Things.

The new generation processor, GrAI VIP, Life-Ready AI Chip, has been increased in capacity compared to its predecessor, making it capable of running larger networks such as ResNet-50, and it has been built on a more advanced processor node. In addition, numerous optimizations have been made to the architecture.

The winners in the 'Hardware' category

Brian Carlson (2nd from left) accepted the embedded award 2022 in the hardware category on behalf of NXP Semiconductor, presented by Peter Ottmann (left), NürnbergMesse, Yasmine Blair (2nd from right), moderator of the award ceremony from NürnbergMesse and Prof. Dr.-Ing. Axel Sikora (right), Head of Jury.

© Uwe Niklas

NXP Semiconductors with the S32 RT Real-Time Processors

NXP's S32 RT real-time processors, based on 16nm FinFET process technology, offer a new class of automotive processing that delivers the benefits of microcontroller behavior but with significantly higher performance, integration and memory expansion. They fill a key processing space between the S32K microcontrollers and the S32G vehicle network processors in the S32 platform to provide the high computing power required for vehicle control, domain control, zonal gateways and safety processing applications.

The S32 RT real-time processor product line offers Automitive developers a scalable price/performance ratio with the S32Z2 and S32E2 series available now.

The winners in the 'Safety & Security' category

Three of CodeClinic attended the award ceremony for the embedded award 2022: (from left to right) Tom Keeble, Isaac Langmead and Neil Langmead, all from CodeClinic, received the award with trophy and certificate in the Safety & Security category from Yasmine Blair (right), moderator from NürnbergMesse, and Prof. Dr.-Ing.

© Uwe Niklas

CodeClinic with the Traci Traceability Solution

CodeClinic's Traci is an innovative traceability matrix that identifies gaps in the development and verification of requirements as required by most traditional and emerging standards such as ISO 22839 and 21434. Traci's traceability engine includes a unique multi-domain visualization of gaps, impact analysis, metrics and reports. The analysis shows the impact of code changes across all domains. The tool provides a unique analysis of the software architecture and its implementation to show how the requirements are connected. Gaps are clearly visible as empty cells in the traceability matrix and are reported as compliance violations.

Traci can be integrated into a CI DevOps pipeline to assess traceability and automatically generate a report demonstrating consistency and completeness for compliance.

The winners in the 'Software' category

STMicroelectronics accepted the award for the embedded award 2022 in the Software category with a team of four: (from left to right) Tarik Souibes, Edouard Dulau, Corinna Babin and Vianney Taufour, surrounded by Yasmine Blair (right), moderator from NürnbergMesse, and Prof. Dr.-Ing.

© Uwe Niklas

STMicroelectronics with the NanoEdge AI Studio

NanoEdge AI Studio from STMicroelectronics is a machine learning library generator that enables any developer to benefit from AI capabilities previously reserved for experts. With this state-of-the-art all-in-one software tool, developers can easily develop an ML-based application for tiny devices based on an STM32 microcontroller. The tool provides an integrated workflow, from data acquisition to inference at the edge, including on-device learning.

In its latest version, NanoEdge AI Studio expands its scope and becomes more universal by integrating support for ISPU MEMS sensors, including the latest ISM330ISN smart sensors with embedded processing unit. AI training and inference can now be distributed across multiple devices, including an MCU and a sensor with an ISPU. This significantly reduces the power consumption of the application. The always-on sensor manages an initial detection phase where AI is performed with very low power consumption and only wakes up the microcontroller when certain events are detected to perform further AI analysis.

The winners in the 'Tools' category

Accemic Technologies received the embedded award 2022 in the Tools category. Dr. Alexander Weiss (left), CEO of Accemic Technologies, and Albert Schulz (right) are delighted to receive the award with trophy and certificate.

© Uwe Niklas

Accemic Technologies with CEDARtools

CEDARtools from Accemic Technologies is the first non-intrusive, continuous live observation tool for embedded systems. The tool runs with released code without influencing it and measures code coverage without instrumentation even at high functional test levels. It can be used to measure complex timing behavior without affecting the system. CEDARtools can process the multi-gigabit trace data stream generated by most industry standard processors such as ARM Cortex-A, Infineon Aurix, NXP QorIQ / Layerscape and Intel Atom via high-speed serial trace interfaces.

It consists of two components:

• the real-time observation of processors based on FPGA-based live trace data processing and
• non-intrusive, continuous live analysis applications when code is released without any instrumentation.

The winners in the 'Startup' category

Vincent Huard (left), CTO of Dolphin Design, and Colin Duggan (right), CEO of BG Networks, are delighted with the award in the Startup category and the associated media budget. No representative from the third award winner, A-China Semiconductor, came to Nuremberg.

© Uwe Niklas

In the Startup category, all three nominees were awarded media packages donated by WEKA Fachmedien:

1. Media package worth 20,000 euros: Dolphin Design with the Edge AI/ML accelerator (NPU) TinyRAPTOR.
   Tiny RAPTOR from Dolphin Design is an energy-efficient NPU-IP (Neural Processing Unit) for TinyML applications. The computing power of the Tiny RAPTOR is based on Processing Elements (PE) with a special instruction set that enables unique near-memory computing. This approach has two direct advantages:
   
   • Both pre-processing and post-processing of neural networks can be managed directly in the accelerator without the need for an external core, providing software flexibility compared to the MAC array approach.
   • Data reuse is maximized as almost no data movement (near-memory) from/to L2 memory is required in TinyML applications, saving a lot of energy at the system level.
     
2. Media package worth 15,000 euros: BG Networks with the Security Automation Tool (BGN-SAT).
   The BGN-SAT Security Automation Tool from BG Networks helps embedded software engineers to quickly improve security, reduce development time and take advantage of in-silicon security features in their IoT developments. With the intuitive, GUI-based software tool, a security profile can be defined to match the cybersecurity requirements of an embedded device. BGN-SAT, complemented by BG Networks' open source security architecture (BGN-ESSA), provides embedded engineers with a low-code software platform to easily develop security controls. They can form the basis for the security of an IoT development, such as secure boot, encryption, authentication and secure software updates.
   
3. Media package worth 10,000 euros: A-China Semiconductor with the MCU007, a 32 bit microcontroller with PUF security function.
   The MCU007 from A-China Semiconductor is a 32-bit microcontroller with PUF security function. The Physical Unclonable Function (PUF) utilizes small differences in the semiconductor manufacturing process that occur naturally and are unpredictable, non-clonable, truly random and highly reliable. A concept comparable to the human fingerprint. The technology is suitable for firmware protection, car keys, IoT authentication, brand protection, licensing, blockchain technology and many other applications with high security requirements such as NFT (Non-Fungible Token) or cryptocurrencies. Device manufacturers can use the MCU007 for new product development and use the PUF ID to protect their firmware. In this way, they can prevent their development from being copied.