“We often see the pursuit of success as being a solitary activity, but Embedded World 2024 was the antithesis of this, with record numbers of attendees and exhibitors coming together to learn about and look toward the future of embedded technology,” said Graham Maggs, VP Marketing EMEA, Mouser Electronics. “Connectivity, AI at the edge, and security were at the forefront of the conversations we had this year.”
Like many others who attended Maggs thought it a much-improved experience on the last few years.
“Embedded World showed a rejuvenated atmosphere and paired with the signs of recovery that we are starting to see in the market, promises to make the second half of 2024 interesting.”
Maggs identified artificial intelligence, along with connectivity, edge computing and security, as the key theme at this year’s show. There were certainly plenty of companies discussing and promoting semiconductors capable of supporting neural network operations and enabling the use of AI in real-world applications such as voice authentication and computer vision.
Among the other trends at this year’s show were sensor fusion to improve the precision and dependability of positioning systems and the growth in what was described as smart embedded technologies – meaning the integration of AI, machine learning and sensors.
Talking to exhibitors and attendees alike other topics coming to the fore included developments in software-defined vehicles and the continued development of ADAS, as well as new innovations across IoT and robotics.
But whatever the discussion AI tended to loom large, and many were looking at using AI to drive growth and to use it to not only become smarter but more efficient.
Intel
Intel ran with that theme and in its event for its partners, held before the show opened, focused on networking, edge computing and AI.
Christine Boles, vice president in Intel's Network and Edge Group and general manager of the Federal and Industrial Solutions division talked at length about software-defined manufacturing and how different types of AI on edge computing were now becoming essential components in industrial systems.
Intel unveiled a number of new chips featuring two new Intel Core processors for the edge, incorporating a GPU and, in the Core Ultra range, a neural processing unit; new Atom processors, as well as an Intel Arc discrete GPU for the edge and Agilex 5 system-on-chip field-programmable gate arrays (FPGAs)
It also took the opportunity to launch the Industry Solution Builders programme - a new 'ecosystem initiative' in the Intel Partner Alliance, which aims to set up communities for specific industries, involving partners and customers from the whole of Intel's partner programme as well as standards bodies to form coalitions to provide open software frameworks, and offer training and support.
AI and Machine Learning
Reza Kazerounian, President and Co-Founder of Alif Semiconductor, focused on the company’s work in terms of endpoint AI and the integration of AI, machine learning and sensors.
"This year we met with numerous European customers, solidifying our position at the forefront of on-device artificial intelligence for embedded systems.
“Alif's capabilities and offerings are gaining traction as designers struggle with traditional GPU-based processors for endpoint AI applications due to limitations in cost, power consumption, and size. Our answers lie in integrated microcontrollers and fusion processors equipped with on-chip neural processing units (NPUs).”
To that end Alif launched the Balletto family at Embedded World.
“This exemplifies our approach. By integrating Bluetooth Low Energy (BLE), Matter protocol support, and an NPU on a single chip, Balletto enables developers to add intelligent audio features like local speech recognition and noise reduction directly on their devices. This eliminates the need for extra components like DSPs (Digital Signal Processors) in space-constrained designs.”
According to Alif these devices are the world’s first Bluetooth Low Energy (BLE) wireless microcontrollers to feature hardware optimisation for AI/ML workloads.
The device’s Arm Cortex-M55 core, which achieves an EEMBC CoreMark score of 704 at 160MHz, includes the Arm Helium M-profile vector extension (MVE), providing a 500% improvement in DSP performance. Alongside the Cortex-M55 CPU, Balletto also includes an Arm Ethos-U55 Neural Processing Unit (NPU), which can perform up to 46 GOPS and is backed by a large 2MB tightly coupled memory (TCM), extending the audio capabilities with hardware acceleration of AI/ML models.
Robotics
BlackBerry used Embedded World to announce a collaboration with AMD that’s intended to help with next-generation robotic systems by enabling lower latency and jitter, and repeatable determinism.
The platform, which was demonstrated on their stand by a new robotic arm, combines BlackBerry QNX expertise in real-time foundational software solutions and the QNX Software Development Platform (SDP) with heterogeneous hardware solutions powered by AMD’s Kria K26 SOM. This combination enables sensor fusion, high-performance data processing, real-time control, industrial networking, and reduced latency in robotic applications.
“This solution from BlackBerry QNX and through our collaboration with AMD will provide an integrated software-hardware foundation that ensures that critical robotic tasks are executed with the same level of precision and responsiveness every single time,” said Grant Courville, VP Product & Strategy at BlackBerry QNX. “These are crucial attributes for industries carrying out finely tuned operations, such as the fast-growing industries of autonomous mobile robots and surgical robotics.”
IoT & Security
Elsewhere, developments in digital security and IoT solutions were announced.
Kudelski IoT revealed that its keySTREAM IoT security product now enables the provisioning of chips from four of the world’s top microcontroller and connectivity IC manufacturers, enabling device OEMs to automate and streamline the provisioning of IoT products at the factory or in the field, simplifying inventory management and onboarding to cloud computing services via a large variety of different chips including secure elements, MCUs and Bluetooth and Wi-Fi solutions.
This provisioning service gives OEMs much greater freedom of choice among leading chipsets while providing a unified provisioning platform. It offers two different models.
The premier integration model is a deep collaboration between Kudelski IoT and the semiconductor manufacturer that embeds the infrastructure for keySTREAM provisioning and lifecycle management within the chipset itself, bundling product and services into a single offering. To that end, Microchip’s ECC608 TrustMANAGER was announced as Kudelski’s first premier integration and includes a self-service root certificate authority and in-field provisioning service powered by keySTREAM as an integrated part of the Microchip product.
The other model is a collaborative partnership and STMicroelectronics, Infineon, and Silicon Laboratories now offer keySTREAM in-field provisioning as an optional service to OEMs.
Commenting Hardy Schmidbauer, Senior Vice President at Kudelski IoT, said, “For many IoT device manufacturers, security provisioning is a laborious manual process that is prone to error. By partnering with top IC manufacturers, we’re making provisioning a simpler, more automated, and secure process than ever before.”
Turning to the IoT, Silicon Labs unveiled its new xG26 family of Wireless SoCs and MCUs consisting of the multiprotocol MG26 SoC, the Bluetooth LE BG26 SoC, and the PG26 MCU that have been designed to future-proof the IoT against emerging applications, like Matter, with double the Flash and RAM of other Silicon Labs multiprotocol devices.
"As users from consumer to industrial sectors extract more benefits from their IoT deployments, their requirements are steadily increasing," said Matt Johnson, CEO of Silicon Labs. "The new xG26 family is built for the future, empowering device manufacturers with the confidence that their current designs will meet tomorrow’s demands."
While the MG26 is intended for Matter other devices in the family have been optimised for Bluetooth LE and Bluetooth Mesh, while the PG26 brings low-power intelligence to unconnected applications like CCTV cameras, remotes, and children's toys.
Astra AI-native IoT Platform
This year’s exhibition also saw Synaptics launch the Synaptics Astra platform along with the SL-Series of embedded AI-native Internet of Things (IoT) processors and the Astra Machina Foundation Series development kit.
The Synaptics Astra AI-native compute platform is intended to help customers who are looking to invest in AI and brings structure to IoT edge device designs using scalable hardware, unified software, an adaptive open-source AI framework, a partner-based ecosystem, and wireless connectivity.
“The SL-Series delivers on the high-performance end of a roadmap of scalable edge AI IoT compute solutions,” said Vikram Gupta, SVP & GM IoT Processors and Chief Product Officer at Synaptics. “Combined with our easy-to-use AI frameworks, however, customers will be able to bring intelligent IoT products to the market far more quickly.”
“Edge AI is transforming use cases across markets including smart home, city, agriculture, and retail,” said Paul Williamson, SVP and GM, IoT Line of Business at Arm. “The Arm-powered Astra platform will deliver the performance and intelligence needed for this next generation of AI experiences.”
Application security testing
TrustInSoft, a Franco-American company, specialising in detailed C/C++ software code analysis - it provides mathematically proven guarantees about software safety and security and significantly reduces software testing time and costs - announced a number of new capabilities for its TrustInSoft Analyzer.
These included the introduction of input coverage metrics for abstract interpretation-based static analysis, as well as enabling tracking of the level of input generalisations when emulating billions of tests.
According to Caroline Guillaume, TrustInSoft CEO, these new capabilities will help to enhance input coverage tracking significantly and will help developers to uncover and rectify gaps in code analysis and achieve unparalleled and accurate test coverage much more easily.
“These advancements in TrustInSoft Analyzer not only enhance our ability to offer mathematical guarantees for zero-defect software but also align with industry standards, particularly for the automotive, IoT, and defence sectors,” said Guillaume.
So there you have it, a brief overview of this year’s Embedded World. It was certainly a very busy show and finally returned to the frenetic atmosphere of shows prior to 2020 and the covid pandemic.
Many of the companies in attendance were focused on edge AI, but there was also considerable interest in power efficiency and delivering more sustainable devices.
Security was also as important as ever, but it was probably software development that seemed to be more apparent at this year’s show.
It’s certainly the case that as systems become more complex so the quality of software will become more important especially when it comes to IoT development.
Today cars, homes, and even clothes are connected to the internet and underscores the importance of ensuring that these devices work as expected.
Companies will certainly need to remember that software is the lifeblood of their IoT products.
