Embedded Systems
14K subscribers
1.07K photos
232 videos
1 file
741 links
News, articles, guides, analytics, projects and startups from the embedded industry.

Topics of the channel:
◽️ machine learning
◽️ internet of things
◽️ embedded software
◽️ advanced electronics
◽️ new engineering ideas

⤵️ Share and subscribe!
Download Telegram
🔐 ESP32 Ad-Blocking Dongle

This project turns a low-cost ESP32-C3 microcontroller into a compact backup DNS ad-blocking device. By using smart data compression techniques, it can block hundreds of thousands of domains while operating with very limited memory and power.

Features of the project:
➡️Blocks 537,000 domains
➡️Uses 40-bit hashing
➡️Requires only 50KB RAM
➡️Offers 10ms DNS responses

The novelty of this project lies in squeezing large-scale DNS filtering into extremely limited hardware. By replacing traditional domain storage with an efficient hashing system, it shows that even a tiny $5 microcontroller can handle tasks usually reserved for more powerful devices.

More information...

#projects #esp32
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
👍9👌6
💻️ Nvidia quietly increases Jetson prices by up to 101%

NVIDIA has raised prices across its Jetson modules and developer kits by up to 101%, following earlier product discontinuations linked to rising LPDDR4 memory costs.

The largest increases affect the Jetson Nano module (from $99 to $199) and the Jetson AGX Orin 32GB module (from $899 to $1,799). The popular Jetson Orin Nano Super Developer Kit has also increased from $249 to $399. The new prices are reflected on NVIDIA's store and most retailers, though some backordered listings may still show the old pricing.

NVIDIA has not provided an official explanation for the increases, but industry observers believe they are driven by rising DRAM memory costs and ongoing supply constraints. The company updated the pricing on its official website without an announcement, leaving it unclear whether the higher prices are a temporary response to market conditions or a permanent change to the Jetson product lineup.

More information...

#news #nvidia
Please open Telegram to view this post
VIEW IN TELEGRAM
👎7👌2
This media is not supported in your browser
VIEW IN TELEGRAM
🧑‍🚀 ESP32-based Plane Radar

Plane Radar is an open-source ESP32 firmware that transforms an ESP32-C3 with a 1.28-inch round display into a live ADS-B aircraft radar. It visualizes nearby aircraft in a sonar-style interface, showing their position, distance and flight details in real time.

Features of the project:
➡️Live ADS-B aircraft tracking
➡️Sonar-style radar display
➡️Flight context and weather
➡️Wireless OTA firmware updates

Plane Radar is a compact and easy-to-build project for aviation enthusiasts and ESP32 makers. The materials are available on ☁️ GitHub. Its combination of real-time aircraft tracking, customizable firmware and convenient wireless updates makes it an appealing choice for anyone interested in embedded hardware or flight monitoring.

More information...

#projects #esp32
Please open Telegram to view this post
VIEW IN TELEGRAM
👍13
This media is not supported in your browser
VIEW IN TELEGRAM
💻 Barely Adequate Browser ESP32

Babe32 is a lightweight text-based web browser for ESP32-S3 touchscreen boards such as the Cheap Black Display. It renders a subset of HTML with 🇳🇱 LVGL library, supports optional image loading, and focuses on practical browsing without JavaScript.

Features of the project:
➡️Direct browsing with proxy fallback
➡️HTML rendering and forms
➡️Touchscreen navigation and keyboard
➡️WiFi, history and power management

Babe32 delivers a surprisingly capable browsing experience on inexpensive ESP32 hardware while remaining lightweight and resource-conscious.

More information:
🔗 The main article
😹 The GitHub repository

#projects #esp32 #lvgl
Please open Telegram to view this post
VIEW IN TELEGRAM
👍9👌2
🧠 Qapla': Training AI on ESP32 microcontroller

Qapla' is an open-source project that demonstrates training a transformer model entirely from scratch on an inexpensive ESP32-S3 microcontroller.

Instead of running a pre-trained model, it performs the complete training process—including backpropagation and weight updates—directly on the device.

Features of the project:
➡️Full on-device transformer training
➡️Handwritten C backpropagation
➡️No pre-trained model required
➡️Open and reproducible setup

Qapla' proves that even low-cost microcontrollers can train small transformer models without GPUs or cloud infrastructure. It is an interesting project for embedded AI developers, TinyML enthusiasts, and researchers exploring on-device learning under extreme hardware constraints.

More information...

#projects #esp32 #artificial_intelligence
Please open Telegram to view this post
VIEW IN TELEGRAM
1👍10👌2👎1
🤖 Impulse - Open-Source Two-Wheeled Bipedal Robot

Developed from the scratch, this project explores a compact robotic platform that combines two-legged articulation with wheeled locomotion. The design prioritizes stability, responsive movement, adjustable height, and portability.

Features of the project:
➡️Self-balancing PID control
➡️Automatic deployment and landing
➡️Variable-height legs
➡️Fully 3D-printed construction

The project combines mechanical design with a compact electronics and control system built around a Teensy 4.1, ODrive motor controllers, magnetic encoders, an IMU, and CAN bus communication. Its 6-DOF structure and multiple PID controllers enable stable balancing, velocity control, and dynamic leg adjustment.

Impulse is particularly interesting for robotics enthusiasts and makers exploring balancing robots, actuator control, and practical control theory. Its detailed hardware and control approach also make it a useful reference for people building their own wheeled robots.

More information:
🔗 The main article
📥 Source files

#projects #robots
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
👍9👌5
📱📱 Samsung Raises Advanced Chip Prices as AI Demand Surges

Samsung has raised prices for new orders on its 4nm, 5nm and 8nm foundry processes, with increases of up to 15% for some customers in China and the U.S., according to Reuters. The biggest hikes are reportedly affecting Chinese chip designers, while Samsung’s 4nm facility in Pyeongtaek has been operating at full capacity since late last year as AI-related demand continues to grow.

The price increases could help Samsung’s loss-making foundry business move toward profitability as demand for advanced chips remains strong and competitors such as TSMC also raise prices. Samsung’s foundry business still trails TSMC significantly in market share, but growing orders from companies including Tesla, Apple, Broadcom and Nvidia are strengthening its position.

More information...

#news #semiconductors
Please open Telegram to view this post
VIEW IN TELEGRAM
👌2👎1
Embedded Systems
🇨🇳 Espressif Unveils ESP32-S31: A Feature-Rich Next-Gen Wireless SoC Espressif announces the ESP32-S31, a new high-performance System-on-Chip designed for advanced IoT devices. The chip combines modern connectivity, processing power, and multimedia support…
🐧 Linux Comes to ESP32-S31

The Chinese company 🇨🇳 Espressif Systems has released a developer preview of its Linux BSP for the 🇳🇱 ESP32-S31, bringing Linux support to the dual-core RISC-V microprocessor. The chip’s MMU and 16MB of PSRAM make running a minimal Linux system possible, while the official Board Support Package (BSP) uses Buildroot and a Linux 6.18-based kernel. Those are available on ☁️ this GitHub repository.

The BSP is currently intended for experimentation rather than production, with limited features, possible breaking changes, and irregular updates. Meanwhile, community developers are also working on Linux ports for the ESP32-S31, with projects exploring newer kernels and features such as Wi-Fi, Bluetooth, and dual-core support.

More information...

#news #esp32 #linux
Please open Telegram to view this post
VIEW IN TELEGRAM
👌5👍4
🤖 Open-Source Humanoid Breaks 400m Record

Robot “Tiangong” made headlines at the 2nd World Humanoid Robot Games in Beijing by completing the 400m race in 38.15 seconds, beating the previous human world record of 43.03 seconds. The robot is part of China’s state-backed X-Humanoid program and is built as a platform for advancing humanoid robotics, with its latest version featuring 43 degrees of freedom and high-dynamic movement control.

A notable part of Tiangong’s development is that some of its technology is open source. X-Humanoid has released components including 😹 locomotion-control software, robot models, training tools and large robotics datasets, allowing researchers and other developers to build on its technology.

More information...

#news #robots
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
👍7👎1👌1