Embedded Systems

📟 Deep-Learning AI on Low-Power Microcontrollers: MNIST Handwriting Recognition Using TensorFlow Lite Micro on Arm Cortex-M Devices Handwriting digit recognition using TensorFlow and MNIST has become a pretty common introduction to artificial intelligence (AI) and machine learning (ML). "MINST" is the Modified National Institute of Standards and Technology database, which contains 70,000 examples of handwritten digits. Today you can create a fully functional MNIST handwriting recognition app on even an 8-bit microcontroller. To demonstrate, we’re going to build a fully functional MNIST handwriting recognition app using TensorFlow Lite to run our AI inference on a low-power STMicroelectronics microcontroller using an Arm Cortex M7-based processor. An implementation like this demonstrates the ability to create robust ML applications even on battery powered, self-contained devices in all sorts edge IoT or handheld scenarios. More information... #projects #machine_learning #cortex #artificial_intelligence

💣 Embedded Game Programming on ESP32 A very interesting set of articles describing the design of a game for Odroid Go, which is based on ESP32. Everything is shown in detail, the code is explained and you will learn a lot of useful tricks from it. The author is a professional embedded developer sharing his experience of software development on a simple example. More information: ▫️ Part 0: Motivation ▫️ Part 1: Build System ▫️ Part 2: Input ▫️ Part 3: Display ▫️ Part 4: Storage ▫️ Part 5: Battery ▫️ Part 6: Audio ▫️ Part 7: Text ▫️ Part 8: Tile System #projects #esp32

📟 GPS/GSM Vehicle Tracker/Recorder The article below is a detailed description of a GPS and GSM based vehicle tracking solution. Its major features: ▫️ Real time location reports using an onboard GSM modem. ▫️ Has a micro SD card socket for offline storage on a memory card – this way using a SIM card is optional. ▫️ Can act as a web server to allow easy and granular customization by the end user. ▫️ No hefty monthly subscription needed for real time or historical data (i.e. where was X car on the 3rd of last month 10:00-13:00). ▫️ Contains no 3rd party trackers. ▫️ All parts of the solution (including the hardware) are open source. More information... #projects #esp32 #gps #IoT

📟 The Hash Monster: ESP32 Tamagotchi For WiFi Cracking A device that saves PCAP: captures into micro sd card to later review them on Wireshark or crack those WPA / WPA2 passphrases. Sounds like something out of the NSA spy WiFi toolset but It's very easy to setup with the ESP32 WiFi Hash Monster and the $35 M5Stack CORE IoT Development Kit. More information... #projects #esp32

🎚 OSHChip (OpenSourceHardwareChip) The most striking thing you see is that it is the size of a 16 pin DIP, and unlike other processor boards, it is designed from the start to be compatible with the white breadboards that are commonly used by electronic hobbyists, robotics projects, and oh so many hackaday projects. Unlike other processor boards that are far larger, OSHChip can be plugged into the middle two rows of a bread board, and it doesn’t cover the connected holes that you will be connecting to. Specifications: ▫️ ARM Cortex-M0 32 bit (16 MHz) ▫️ 2.4 GHz Bluetooth Low Energy radio ▫️ Broad range of built-in Peripheral functions ▫️ 256 KBytes of Flash memory for program and data ▫️ 32 KBytes of SRAM ▫️ Support for in-system debugging including breakpoints More information... #boards #cortex

📟 Automatic Trash Classification with Raspberry Pi and Arm NN Machine Learning AI is now moving to edge devices. Arm is leading this revolution by providing accelerated hardware IPs and middle-layer solutions. In practice, this means you can train your models using the most popular ML frameworks. You can then load models and optimize them on your edge device to perform fast inference, eliminating network and privacy problems. To demonstrate ML with edge computing resources, we'll use Arm NN APIs to classify images of trash from a webcam attached to a Raspberry Pi. The Raspberry Pi will display classification results. More information... #projects #raspberry #artificial_intelligence #machine_learning

📰 Blue pill VS Black pill: Transitioning from STM32F103 to STM32F411 For many years now, the so-called ‘Blue Pill’ STM32 MCU development board has been a staple in the hobbyist community. Starting last year, boards with a black soldermask and an STM32F4 Access (entry-level) series MCUs including the F401 and F411 began to appear. These boards with the nickname ‘Black Pill’ or ‘Black Pill 2’. F103 boards also existed with black soldermask for a while, so it’s confusing. The F4xx Black Pills are available via the same sources as the F103-based Blue Pill ones, for a similar price, but feature an MCU that’s considerably newer and more powerful. This raises the question of whether it makes sense at this point to switch to these new boards. More information... #news #stm32 #hardware

📟 Interesting experiment: Making cheap IR night camera out of ESP32-CAM by removing IR Filter Result after removing IR filter (part 2): Having the IR confined to an area where it can bounce around and not just be lost to the night makes a big difference. With a better (or more than one) source of IR, the ESP32-CAM may well make a decent indoor night camera. Tweaking some of the camera settings may improve things, too. More information: ▫️ Part one... ▫️ Part two... ▫️ Part three... #projects #esp32 #experiments #camera

🎛 Dog Mood Detector with Google Cloud and Raspberry Pi In this article we’re going to explore the idea of a dog mood detector. We’ll build a device that listens to the ambient sound around it and, if a dog is present, attempts to determine which type of sound the dog is making: a friendly bark, a scared whimper, or an aggressive growl. Based on the preferences of the user, the device vibrates when it thinks they should check on their dog. This could be used, perhaps, to help owners keep tabs on their dog when they're out of earshot. To prototype this device, we’ll use a Raspberry Pi, which is my go-to platform for implementing Machine Learning on edge devices. More information... #projects #raspberry #machine_learning #IoT #python

📟 Creating a Text-to-speech engine with Google Tesseract and Arm NN on Raspberry Pi A text-to-speech (TTS) engine is the crucial element of systems looking to create a natural interaction between humans and machines based on embedded devices. An embedded device can, for example, help visually impaired people read signs, letters, and documents. More specifically, a device can use optical character recognition to let a user know what it sees in an image. In this article, you will learn how to do so with TensorFlow, OpenCV, Festival, and a Raspberry Pi. I will use the TensorFlow machine learning platform along with a pre-trained Keras-OCR model to perform OCR. OpenCV will be used to capture images from the web camera. Finally, the Festival speech synthesis system will serve as the TTS module. Everything will then be combined to build the Python app for Raspberry Pi. More information... #projects #raspberry #machine_learning #artificial_intelligence #python

🤖 Simple Telegram Bot with ESP32: Control GPIO Pins through Telegram Chat In this tutorial, we will be controlling a LED with the Telegram bot, the LED is connected to the ESP32 board. Instead of the LED, you can control any pin connected to any other appliance or component. Telegram allows us to create several bots with different functionalities. For our project, we will be creating a simple bot using Telegram. The rest of the command declarations and replies will be coded in the ESP board itself, which will communicate to our bot using the chat ID. More information... #projects #IoT #esp32 #bot

📰 Pine64 Announces 'Next Generation' Quartz64 Single Board Computer Pine64 has announced the Quartz64 model-A AM SBC, which seems to have an impressive set of features: Powered by a Rockchip RK3566 SoC, the Quartz64 comes with a quad-core Arm Cortex A55 CPU running up to 1.8GHz and a Mali-G52 GPU running at 800 MHz which supports Vulkan 1.0 and 1.1. Quartz64 supports up to 8GB of LPDDR4 RAM, but there will be models starting from 2GB of RAM. Quartz64 model-A is quite a large board, larger than a Raspberry Pi 4 for sure. Measuring 5.2 x 3.1 x 0.7 inches (133 x 80 x 19mm), the Quartz64 offers plenty of extra space for extra features and a port layout that takes advantage of all the extra space. More information... #news #boards #cortex #hardware

📰 Alternatives to Raspberry Pi in 2021: ▫️Hackboard 2 is a new single-board computer designed from the ground up to be one of the most powerful and affordable Windows-capable, Intel-based computers ever. With optional 4G or 5G connectivity, it is ideal for students, teachers, families, makers, and hobbyists alike. Unlike the Raspberry Pi, Hackboard 2 is based on a powerful Intel processor capable of running Windows. ▫️ZimaBoard is a low-cost single board server designed for makers and geeks. Zimaboard has both the expandability of an x86 SBC, and the appropriate power of a micro server. With ZimaBoard, you can setup a 4 Terabyte personal cloud in 5 minutes, configure a secure VPN to protect your digital footprint, tune your network with OpenWrt and pfSense, build a 4K media server that runs with Plex, share and collaborate with team members on your files, or upgrade your smart projects across different systems. ▫️ROCK Pi 4 is a Single Board Computer in an ultra-small form factor that offers class-leading performance while leveraging outstanding mechanical compatibility. The ROCK Pi 4 offers makers, IoT enthusiasts, hobbyists, PC DIY enthusiasts and others a reliable and extremely capable platform for building and tinkering their ideas into reality. ▫️ODROID-H2+ features an Intel Celeron J4115 Gemini Lake processor, two SO-DIMM slots for memory, two SATA ports and M.2 NVMe slot for storage, two Gigabit Ethernet ports, a mix of USB 3.0 and 2.0 ports, and more. This very powerful x86 64bit single board computer has large memory capacity and advanced IO ports. ▫️The NanoPi R4S (as "R4S") is an open source platform with dual-Gbps Ethernet ports designed and developed by FriendlyElec for IoT applications. The NanoPi R4S uses the RK3399 SoC. It has two Gbps Ethernet ports and 1G/4G DDR4 RAM. FriendlyElec ported an OpenWrt system for it. It works with Docker CE. It is a good platform for developing IoT applications, NAS applications, smart home gateways etc. ▫️The NVIDIA® Jetson Nano™ 2GB Developer Kit is ideal for learning, building, and teaching AI and robotics—built for creators and priced for everyone. The Jetson Nano 2GB Developer Kit delivers incredible AI performance at a low price. It makes the world of AI and robotics accessible to everyone with the exact same software and tools used to create breakthrough AI products across all industries. ▫️ASUS Tinker Board is a Single Board Computer in an ultra-small form factor that offers class-leading performance while leveraging outstanding mechanical compatibility. The Tinker Board offers makers, IoT enthusiasts, hobbyists, PC DIY enthusiasts and others a reliable and extremely capable platform for building and tinkering their ideas into reality. #news #boards #raspberry #hardware #jetson_nano

🎛 Smart Video Doorbell: Intercom on Raspberry Pi I came across some nice software to make a fast video / 2-way audio doorbell with a Raspberry Pi, and found a way to fit all the components in a small case. I will show you how to build it, including how to make push notifications work when you press the button. More information... #projects #raspberry #IoT

🎛 Machine Learning Based Gesture Detection Watch (ESP8266) We’ll be taking the watch built in the previous instructable and plunge in some Machine Learning and see if we can detect what kind of gesture we are performing, and maybe in upcoming instructable we can use this gesture to perform some very cool projects based on this. Final result: Once the code is uploaded successfully make any gesture, the oled display should say what gesture you are performing. In my case the model worked 95% of the time, it just sometimes it suffers to find the right swipe gesture. Probably, the data I collected was noisy or maybe I wasn’t doing the motion right while collecting the data. Anyhow 95% is still good for me, we can do so much with this gesture detection! More information... #projects #machine_learning

📟 New development board ESP32-C3 ESP32-C3 attempts to address the most common needs for connected devices. Here are the key criteria that have been considered for the design of ESP32-C3: ▫️ Security is of prime importance. Even the lowest-cost connectivity solution needs to provide an appropriate level of security for common security threats. ▫️ Bluetooth Low Energy availability is useful for improving user experience and field diagnostics. ▫️ While cost is a very important parameter, the availability of sufficient memory for common use-cases is equally important. ESP32-C3 is a single-core, 32-bit, RISC-V-based MCU with 400KB of SRAM, which is capable of running at 160MHz. It has integrated 2.4 GHz Wi-Fi and Bluetooth LE 5.0 with a long-range support. It has 22 programmable GPIOs with support for ADC, SPI, UART, I2C, I2S, RMT, TWAI, and PWM. More information... #boards #news #esp32

📟 Arduino PRO: PORTENTA H7 The promo video of a new pro development board presented on CES 2020. H7's main processor is the dual core STM32H747 including a Cortex®M7 running at 480 MHz and a Cortex® M4 running at 240 MHz. Use Portenta when performance is key, among other cases, we envision it to be part of: ▫️ High-end industrial machinery ▫️ Laboratory equipment ▫️ Computer vision ▫️ PLCs ▫️ Industry-ready user interfaces ▫️ Robotics controller ▫️ Mission-critical devices ▫️ Dedicated stationary computer ▫️ High-speed booting computation (ms) More information... #news #video #boards #arduino #python #stm32 #cortex