GRINN’s Post

Edge AI platform with Lorandt Fölkel and Adithya Madanahalli of Würth Elektronik Group at embedded world Exhibition&Conference Embedded World 2026 • 🤖 Edge AI at the device level: local inference for computer vision and sensor data, without relying on the cloud • 🧠 Demo built around the GRINN GenioBoard Edge AI SBC / GenioSOM-700 platform • 📷 Good fit for face recognition, object detection, robotics perception, and industrial IoT use cases • 📡 Flexible connectivity with LTE, Wi-Fi, Bluetooth LE, Ethernet, camera inputs, and M.2 expansion • 🔌 More than a board demo: Würth also highlights EMC, power design, thermal, RF, antenna matching, and compliance support • 🏭 Interesting direction for companies building embedded Linux and edge AI products that need a reference design and a path toward certification The GenioBoard Edge AI SBC is positioned by Würth Elektronik and Grinn as an edge AI platform for IoT and computer vision, built around the GenioSOM-700 with a MediaTek platform, 4.0 TOPS AI performance, HDMI/DisplayPort, M.2 expansion, and wireless module options. #EmbeddedWorld2026 #WurthElektronik #EdgeAI #EmbeddedSystems #ComputerVision #IoT #Robotics #IndustrialIoT #Linux #Mediatek You can also watch this video in 4K60 over at my YouTube: https://lnkd.in/dAVXVsay

Back at embedded world 2026! 🚀 It was a pleasure to present our Edge AI platform alongside Lorandt Fölkel. A big thanks to Nicolas Charbonnier for stopping by our booth to film this deep dive. We discussed how we’re making local inference—without the cloud—a reality for industrial vision and sensor applications. Beyond the 4.0 TOPS AI performance, we’re focusing heavily on the "connected" side of the GenioBoard: 📡 Wireless Connectivity: Seamless integration of LTE, Wi-Fi, and BLE. 📡 RF design: Highlighting the critical role of antenna matching and EMC for robust #IoT designs. Check out the full 4K60 video by Nicolas here: https://lnkd.in/dAVXVsay

🚀 Day 4: Bridging the Gap Between Analog Signals and the Cloud! I’m thrilled to share my progress from Day 4 of my Embedded C and IoT journey. Today was all about data—how we capture it from the physical world and how we visualize it in real-time. 🔌 Morning Session: The Power of ADC We kicked off the day by diving deep into Analog-to-Digital Conversion (ADC). The Theory: Understanding how continuous analog signals (like temperature or light) are converted into discrete digital values that a microcontroller can process. The Practice: I utilized Wokwi to simulate sensors like the LDR (Light Dependent Resistor) and DHT22, observing how varying physical inputs translate into digital data. 🌐 Afternoon Session: Entering the World of IoT In the afternoon, we leveled up to IoT Basics, integrating our hardware simulations with the cloud using the Blynk platform. Key Projects Completed: Remote LED Control: Successfully implemented a "Hello World" of IoT by toggling an LED on my simulation from a mobile dashboard. Distance Monitoring: Used an Ultrasonic Sensor to measure distance and pushed that data to the Blynk cloud for real-time tracking. Safety First with Gas Sensors: Integrated a Gas Sensor simulation to monitor air quality levels, visualizing the data through gauges on the Blynk app. 🛠️ Tech Stack: Language: Embedded C Simulation: Wokwi IoT Platform: Blynk Hardware Focus: ESP32, LDR, Ultrasonic Sensor, Gas Sensor It’s amazing to see how a few lines of code can connect a virtual sensor to a mobile device across the internet. Looking forward to what Day 5 brings! #EmbeddedSystems #IoT #Blynk #Wokwi #EmbeddedC #ESP32 #LearningJourney #TechInnovation #Engineering

🚀 First Experiment with MPU6050 + ESP32 Today I interfaced the MPU6050 motion sensor with ESP32 for the first time and created a small motion-controlled game on my laptop. 🔧 Hardware Setup • ESP32 microcontroller • MPU6050 IMU sensor • I2C communication (SDA → GPIO 27, SCL → GPIO 26) 📊 What I Did • Read accelerometer data from MPU6050 • Sent sensor values to the laptop using serial communication • Used Processing to visualize the motion and control a bird game using sensor tilt 🎮 Result Tilting the MPU6050 moves the bird in the game. This helped me understand how motion sensors can control interactive applications. This was a great beginner experiment to learn: • Sensor interfacing • I2C communication • Real-time data visualization • Motion-controlled systems Looking forward to building more Embedded + IoT projects with ESP32. #ESP32 #MPU6050 #EmbeddedSystems #IoT #Electronics #SensorInterfacing #EngineeringProjects

Excited to share the capstone project from the last semester (3rd year even semester, EEE, RUET) TinyML-Based Intelligent Automatic Circuit Recloser (ACR) for power distribution systems This project combines Signal Processing + Embedded AI + IoT to enable real-time fault detection and monitoring on low-cost hardware. Key Highlights: • Real-time fault classification on STM32 (TinyML) • 99.94% classification accuracy • Fault distance estimation with ~1.21% error • Millisecond-level inference latency • IoT-based monitoring using ESP32 • Fully validated on hardware prototype core (STM32) Why it matters: Traditional ACR systems rely on fixed thresholds and lack adaptability. Our solution brings intelligence directly to edge devices, enabling faster and smarter protection. Tech Stack: STM32F446RE, X-CUBE-AI, ESP32, Signal Processing, TinyML #EmbeddedAI #TinyML #IoT #PowerSystems #STM32 #EdgeAI #EngineeringProjects

🔧 Smaller Core, Smarter Integration For developers working on thermal-enabled devices, integration is often the real challenge—not just imaging. The DyMN Series Thermal Imaging Core is designed to simplify that process while keeping performance stable and power consumption low. ⚙️ Built with a patented Falcon image processing chip, it replaces traditional FPGA solutions—helping reduce size, cost, and system complexity. Why it works for embedded applications: • Compact design (21×21mm, lightweight) for easy integration • Low power consumption (<0.65W) for portable or battery systems  • Flexible interfaces (SPI/DVP, I2C) for faster development • Up to 640×512 resolution with clear thermal output • Wide temperature measurement for industrial and smart device use 💡 Integration tip: Choosing a low-power thermal core can significantly extend runtime in UAVs, handheld devices, or IoT systems—without sacrificing imaging quality. Looking to integrate thermal imaging into your product? Let’s talk. 👇 🌐 www.dyt-ir.com 📩 dianytech@gmail.com #dianyangtechnology #thermalimaging #thermalcore #embeddedvision #infraredsensor #iotdevices #machinevision #productdevelopment #smartdevices #engineering

💡 Is the ESP32 the ultimate IoT workhorse? The ESP32 SoC (System-on-Chip) from Espressif Systems has completely transformed embedded systems development. Why? Because it offers an unparalleled balance of high-performance processing, advanced connectivity, and accessibility. This detailed infographic breaks down exactly why this powerful little chip is the default choice for engineers and developers worldwide. Here’s why it dominates the field: Key Engineering Features: ⚡ Dual-Core Processor: Running Tensilica Xtensa LX6 processors, capable of real-time multitasking. 🌐 Native Connectivity:* Seamless Wi-Fi and Bluetooth integration are built-in. 🛠️ Versatile Interfaces: Multiple serial interfaces, including SPI, I2C, UART, and CAN. 🎛️ Rich Peripherals: Advanced control with onboard ADC, DAC, and PWM for sensors. 🔋 Low-Power Efficiency: Dedicated modes designed for battery-powered IoT applications. It is powering innovation across numerous sectors, including: 🏡 Smart Home Systems 📡 Wireless Sensor Networks 🏭 Industrial Monitoring 🤖 Robotics & Automation 📊 Data Acquisition Systems For engineers, researchers, and developers, the ESP32 provides the ideal platform to prototype rapidly and deploy connected solutions efficiently. Are you using ESP32 in your current projects? What is the most innovative application you've built with it? Let me know in the comments! 👇 #ESP32 #IoT #EmbeddedSystems #Electronics #ElectricalEngineering #Microcontrollers #Innovation #TechDevelopment #EspressifSystems

I recently developed a Smart Motion Detection & Alert System using ESP32-S3 and ESP-IDF. This project started as a basic camera streaming server, but I expanded it into a complete IoT-based surveillance solution. 🔧 Key features: 🎥 Live MJPEG video streaming via built-in web server 🌐 Wi-Fi connectivity for local monitoring 🚨 PIR sensor for real-time motion detection (interrupt-based) 📸 Automatic image capture when motion is detected 📲 Instant Telegram alerts with captured images 🧠 From an engineering perspective, this project involved: Embedded HTTP server implementation Camera driver configuration and optimization FreeRTOS task management for event-driven design HTTP client communication with Telegram Bot API Memory handling for image processing on ESP32 One of the main challenges was managing limited resources while handling real-time image capture and network communication efficiently. This project strengthened my practical skills in embedded systems, IoT, and firmware development, and showed how to turn a simple prototype into a real-world smart system. Next step: integrating edge AI (object detection) to make the system more intelligent. #ESP32 #IoT #EmbeddedSystems #SmartSecurity #ESP32S3 #Firmware #EngineeringStudent #EdgeComputing

Bringing GPU Acceleration to Embedded Systems with CUDA C++ Using GPU acceleration from NVIDIA platforms, I optimized real-time workloads such as signal processing and data filtering—reducing latency significantly compared to traditional CPU-only embedded designs. 🔹 Improved real-time performance 🔹 Efficient parallel task execution 🔹 Scalable architecture for AI-ready edge devices This approach opens new possibilities for high-performance embedded applications in areas like smart monitoring, industrial automation, and intelligent IoT systems. Looking forward to refining this further and exploring real-world deployments. #EmbeddedSystems #CUDACPP #EdgeComputing #ParallelProcessing #Innovation

Embedded World 2026. Good show! Security and CRA compliance were the dominant themes, alongside AI at the edge, machine vision, FPGA, and low power. Military presence was minimal due to dual-use concerns. Automotive quieter than expected. No major surprises, but solid technology across the board. The most notable shift: security is no longer a checkbox. It's becoming a starting point. It was front and centre, and vendors who used to lead with performance are now leading with compliance. That's a genuine shift. Good to see NXP Semiconductors STMicroelectronics Analog Devices XMOS Semtech BlackBerry QNX Direct Insight Ltd and many others. Also great to catch up with Paige Hookway at Electronic Specifier Ltd. More in the video on the show's key themes. What about you - what caught your eye at the embedded world Exhibition&Conference ? #EmbeddedWorld #Security #IoT #EdgeAI