Join us as we revisit Ultralytics' experience at Embedded World 2025, where we got to see Ultralytics YOLO models in action across various demos.
Embedded World 2025 in Nuremberg took place from March 11th to 13th, bringing together some of the brightest minds in embedded systems and AI. It was a great opportunity for Ultralytics to showcase our latest Vision AI solutions and demonstrate how our Ultralytics YOLO models are shaping the future of embedded AI.
Throughout the event, we presented live demos of our models, letting attendees see firsthand the capabilities and versatility of our computer vision technology. It was exciting to engage with engineers, manufacturers, and AI enthusiasts from all over the world, exploring new applications and discussing the role of AI in embedded systems.
Embedded World 2025 also gave us the chance to connect with industry leaders and gather valuable feedback, helping us better understand the needs and challenges within the embedded tech space. From learning about new hardware platforms to discussing the changing trends in AI, the event provided plenty of opportunities for collaboration and growth.
In this article, we’ll take you through our experience at Embedded World 2025, including highlights from the demos, insightful discussions, and the exciting future of embedded AI. Let’s get started!
Embedded World is a global event for the embedded systems industry, held annually at the Exhibition Centre in Nuremberg, Germany. Embedded systems are specialized computing devices designed to perform specific tasks within larger systems, often with real-time computing constraints.
Since its foundation in 2003, the event has grown from two halls and around 350 exhibitors to become the leading platform for the embedded community. Over the past 22 years, it has expanded significantly, with over 1,100 exhibitors showcasing their products to more than 32,000 visitors across seven halls in 2024.
The event attracts a diverse group of experts, exhibitors, and attendees, all focused on exploring advancements in embedded systems, hardware, software, services, safety, and more.
Interestingly, the event’s global influence doesn’t stop in Germany. Embedded World has now expanded internationally, with regional events like Embedded World China, taking place in Shanghai from June 11 to 13, 2025, and Embedded World North America, which will be held from November 4 to 6, 2025, in Anaheim, California. The expansion of Embedded World into new regions is an indirect result of the global demand for embedded AI solutions.
Looking specifically at this year's event, Embedded World 2025 hosted over 1,200 exhibitors from around the world. There was a strong focus on embedded AI and computer vision, with a particular emphasis on embedded vision. Embedded vision refers to the integration of vision systems (such as cameras or sensors) into embedded devices, enabling them to process and interpret visual data in real-time. This technology is essential in applications like object detection, facial recognition, and autonomous navigation, which are becoming rapidly adopted in industries ranging from automotive to healthcare.
The Embedded Award 2025 recognized nine winners, many of whom were working on innovations in these areas, highlighting their growing importance in the industry.
Prof. Dr. Axel Sikora, Chairman of the Embedded World Conference and head of the Embedded Award jury, noted, “It is particularly interesting to see that in most cases, it is more than one company working in the same direction, as can be seen from the developments in the field of embedded AI or embedded vision in particular.”
Day one at Embedded World 2025 made it very clear that YOLO has been making waves in the embedded AI space. Our computer vision models, Ultralytics YOLOv8 and Ultralytics YOLO11, are being used as the industry standard for testing AI neural networks on embedded hardware. Everywhere we looked, manufacturers were leveraging YOLO to showcase their latest innovations, from tiny edge AI chips to high-performance computing boards.
In fact, there was a widespread adoption of YOLO by board manufacturers. Whether it was for object detection through bounding boxes or pose estimation, YOLO-powered applications were a dominant theme throughout the exhibition. The robustness, efficiency, and versatility of YOLO make it the go-to choice for embedded AI applications.
Attendees also had the chance to see how YOLO performs across a variety of hardware architectures. From microcontrollers to FPGAs (Field-Programmable Gate Arrays) and full-fledged CPUs (Central Processing Units), our YOLO models demonstrated impressive performance across different systems.
A particularly interesting showcase on day two of the event came from D-Robotics, who ran YOLO on their compact RDKX5 board at an incredible 10ms per frame. This was achieved while maintaining excellent thermal performance, highlighting the efficiency and optimization of YOLO for real-time embedded applications.
Beyond seeing YOLO running on various hardware, we also noticed its growing impact on edge AI platforms. For example, Infineon Technologies featured its DEEPCRAFT™ Studio, which now supports computer vision leveraging Ultralytics YOLO models. This platform allows developers to create low-power, efficient Edge AI models for vision-based applications.
One of the best parts of Embedded World 2025 was the chance to meet directly with the hardware community in Nuremberg. Speaking with engineers, developers, and industry leaders gave us valuable insights into how YOLO is being used in embedded systems, from edge AI chips to high-performance computing boards. These conversations helped us better understand the challenges the industry faces, such as balancing power efficiency with performance and ensuring compatibility with a variety of hardware.
The feedback we received was very valuable, especially when it came to optimizing YOLO for low-power devices and improving deployment across different chipsets. Many expressed their opinions on how the collaboration between hardware and software is driving the next generation of embedded AI solutions. Based on the insights gathered, we’re excited to continue refining our models and supporting the growing demand for smarter, more efficient technology in embedded systems.
Another highlight of Embedded World 2025 was the incredible sense of collaboration. The event gave us a chance to connect with some of the key players in the embedded AI space, opening up exciting opportunities for the future.
For instance, we had the opportunity to engage with STMicroelectronics, who exhibited a fantastic Ultralytics YOLO demo at their booth, demonstrating how our models power their AI applications. We were also featured on STMicroelectronics' Wall of Fame.
.webp)
On top of that, Seed Studio featured their reCamera, which integrates YOLO for real-time vision applications. Similarly, M5Stack displayed their beautifully designed products, all running YOLO-powered AI. It was inspiring to see YOLO being integrated into so many cutting-edge solutions.
Our time at Embedded World 2025 reinforced one central theme: YOLO is leading the way for the next generation of embedded AI solutions. As more industries embrace edge AI, Ultralytics YOLO models continue to deliver the performance, flexibility, and ease of integration that developers and manufacturers need to build innovative products.
As a matter of fact, the embedded AI market is growing rapidly. In 2023, it was valued at $9.13 billion and is projected to reach $29.26 billion by 2032. This estimation hints at the increasing importance of edge AI in shaping the future of technology.
We’re looking forward to continuing to work with industry leaders to expand the horizons of what AI can achieve at the edge. As the growth of intelligent vision applications and new hardware advances, the future of embedded AI looks incredibly promising.
Embedded World 2025 was an exciting event, and Ultralytics had a great experience being part of it. We had the chance to connect with experts, showcase our Vision AI solutions, and learn more about the future of embedded systems.
From live demos featuring Ultralytics YOLO models to exploring the role of embedded AI, the event was full of interesting takeaways. We want to extend a huge thank you to everyone who stopped by to chat with us, share their experiences, and showcase their amazing work using YOLO. Your insights and enthusiasm were invaluable, and we’re grateful for the opportunity to connect.
We’re excited to continue advancing AI technology and are eager to see what’s next in computer vision. Stay tuned for more updates as we continue to grow and innovate.
Join our vibrant community! Dive into our GitHub repository to learn more about AI. If you're looking to incorporate computer vision into your organization, check out our licensing options. Explore the benefits of computer vision in healthcare and see how AI in manufacturing is making a difference!
Begin your journey with the future of machine learning
