Uladzislau Bayouski

  In this episode of Bots, Boards & Brainstorms, we dive into the latest breakthroughs from Google I/O 2025 — including Aloha 2, the open-source robot powered by Gemini AI, and the surprisingly powerful MuJoCo simulator (yes, you can control Spot from your laptop). We break down what it really takes for robots to “see” and “act,” with new research on smarter vision sensors and motion planning using diffusion models. Plus, an autonomous drone that learns in-flight and a shoutout to some wild PCB designs — like a DIY cat feeder and a wearable GPS badge. Whether you're an AI enthusiast, hardware tinkerer, or just want to know how robots are getting smarter (and cooler), this episode delivers sharp insights with zero fluff. Aloha 2 – Stanford's Dual-Arm Robot Platform MuJoCo – Physics Engine for Robotics Simulation “Just Give Me the Outline” – AI Planning for Robot Manipulation (MIT) MIT: Robots Solve Manipulation Tasks in Seconds MIT: AI Control System Stabilizes Drones in Un...

Imagine your basil sending you a message: "Hey, I'm parched." Or your tomatoes flagging a humidity drop before you even walk into the greenhouse. That's not science fiction. That's a Raspberry Pi, a soil moisture sensor, and about a weekend of building. 🌱 The Short Version A DIY smart greenhouse is one of the most satisfying IoT builds you can do — genuinely useful, surprisingly affordable, and a brilliant excuse to learn automation, computer vision, and data visualisation all at once. The core hardware stack: DHT22 — temperature and humidity monitoring Soil moisture sensors — know exactly when your plants need water, not just when you remember to check Relay + water pump — automatic watering triggered by sensor thresholds, not guesswork 💧 Light sensors + LED strips — extend daylight hours and compensate for cloudy days automatically Raspberry Pi or ESP32 — the brain, running ESPHome, Home Assistant, or a lightweight Flask app Pi Camera or webcam ...

Both are under $10. Both have Wi-Fi. Both have dual cores and thriving communities. Both are sitting in the parts drawers of every serious IoT maker on the planet. So which one do you actually reach for? The Short Version The ESP32 and the Raspberry Pi Pico W are the two microcontrollers that dominate budget IoT — but they have genuinely different personalities, and the right choice depends entirely on what you're building. ESP32 is the veteran. Feature-packed, battle-tested, and wireless-first: Dual-core Xtensa processor with FreeRTOS baked in Built-in Wi-Fi and Bluetooth (classic + BLE) 📶 30+ GPIOs with analog input, PWM, I2C, SPI, UART Deep sleep modes that stretch battery life for years Arduino, ESP-IDF, PlatformIO, MicroPython — pick your stack Best for: smart home controllers, wireless sensor nodes, wearables, anything needing Bluetooth or real-time RTOS tasks. Raspberry Pi Pico W is the precise one. Efficient, predictable, Python-friendly: Dual-core ARM Cortex-M0+...

You've written your code. You've flashed it to the board. The LED blinked. Victory! And then… nothing. Welcome to IoT debugging — where things almost work, and then don't. This moment isn't failure. It's your first clue. 🔦 The Short Version Every IoT maker hits the post-blink void. Here's the checklist that gets you out of it: 1. Open the serial monitor first. Insert Serial.println("Here!") debug prints to track where your code gets stuck. Don't forget Serial.begin() and the correct baud rate — without those, your messages disappear into the void. More than half of "mysterious" failures are solved here. 2. Check the hardware before touching the code. Loose jumper wires. Wrong GPIO pin declared. Sensor not getting enough voltage. These cause the majority of IoT failures. Many sensors won't run on 3.3V logic, and some draw more current than a microcontroller pin can supply — add a separate power source with common ground and watch ...

In this episode, we explore projects and products that move intelligence closer to the action—from backyard greenhouses to wildfire zones. In this episode: - ESPHome 2025.5.0 introduces Packet Transport for UART/UDP device-to-device comms—perfect for farm-scale automation with no Wi-Fi stress. https://esphome.io/changelog/2025.5.0.html - A LoRa-based soil monitoring project using Raspberry Pi Pico, solar power, and Home Assistant. https://www.briandorey.com/post/pi-pico-lora-remote-soil-monitor - How a DIY Everything Remote simplified an overcomplicated projector setup—made with ESP32, ESPHome, and a 3D-printed case. https://www.thestockpot.net/videos/theeverythingremote - Meet The Guardian, a fire-seeking robot with YOLO-based flame detection, built to help battle wildfires autonomously. https://blog.adafruit.com/2025/05/25/the-guardian-presenting-a-robot-for-wildland-firefighting/ - New launches: Homey Pro mini and SmartThings developer tools aimed at simplifying smart home integrat...

Your coffee machine knows when you wake up. Your TV knows what you watch and for how long. Your robot vacuum has mapped every room in your home. None of them asked permission. None of them told you where that data goes. The Short Version The "smart" in smart home comes with a side effect most manufacturers don't advertise: your devices are constantly collecting behavioural data and transmitting it upstream. Not maliciously — it's just how the business model works. Usage telemetry, preference data, and behavioural patterns are valuable. Your appliances are part of that pipeline whether you know it or not. Here's what's actually talking: Smart TVs — viewing habits, content preferences, watch duration, and ad engagement sent to manufacturers and third-party data brokers. ACR (Automatic Content Recognition) identifies exactly what's on screen, including content from HDMI inputs 📺 Coffee machines & kitchen appliances — brew schedules, usage frequency, ...

🎥 Exploring Japan’s Advanced Tech Playground – Inside the TEPIA Advanced Technology Gallery In this immersive walkthrough, we visit the TEPIA Advanced Technology Gallery in Tokyo, a hidden gem where science, engineering, and creativity converge. From personal flying machines and 3D printing to robotic assistants and hands-on tech workshops, this experience captures Japan’s cutting-edge innovation—up close and interactive. 🧠 What’s Inside: A sneak peek at futuristic prototypes like SkyDrive and assistive robots Medical tech, expression cameras, and energy-harvesting systems AI-enabled robots with modular designs and 360° maneuvering capabilities Smart traffic systems with person recognition and adaptive lighting Interactive learning stations for kids and families Real-time laser cutting, soldering, 3D printing demos, and hands-on build spaces ✨ Highlights: Creative Lab (2nd Floor): Fully equipped for DIY robotics, IoT, and coding Student-led innovations in robotics, automation, and ...