Using IoT and AI In this project, I designed and built a small-scale vertical-axis wind turbine with IoT sensors to collect real-time environmental data. Using statistical analysis and machine learning, I explored how wind speed and environmental conditions affect power generation and built an AI model to forecast future trends.
I’ve always been curious about how renewable energy works, so I decided to make my own small wind turbine and build a power generation forecasting system around it. Instead of just reading about it, I wanted to actually see it in action — measuring wind, collecting data, uploading everything to the cloud, and process it using AI to forecast.
To start, I 3D printed the blades and crafted a metal stick to hold them together as a vertical wind turbine. I added a rotary encoder to measure how fast it spins, and I connected environmental sensors that record temperature, humidity, and altitude. An ESP32 sends all the data to the cloud, while a small LCD screen shows the turbine’s status in real time. This is the 49-second timelapse of me building it.
timelapse.movWarning: this section gets a little technical — but that’s the fun part!
Disclaimer: this is just an estimation, not a lab-grade measurement
To make my turbine more than just spinning blades, I used math and sensors to measure how much energy it can actually produce. Here’s the step-by-step process:
1. From Pulses to RPM
The encoder creates tiny electrical pulses every time the turbine spins. By counting how many pulses happen in one second, and dividing by the number of pulses per full spin (in my case, 1200), I know how many rotations per second the turbine makes. Multiply that by 60, and I get RPM (rotations per minute).
2. From RPM to Wind Speed
Each full spin of the blade covers the circle’s circumference. Circumference = pi × diameter.
So, wind speed = revolutions per second × circumference.
For my turbine with a 0.33 m diameter, one spin equals about 1.04 meters of air moved.
3. From Wind Speed to Power
Wind carries kinetic energy, and the power in the wind can be estimated with this formula:
Power = 0.5 × air density × swept area × (wind speed³) × efficiency factor
Where:
Air density ≈ 1.225 kg per cubic meter (at sea level)
Swept area = pi × (diameter / 2)²
Wind speed = calculated from RPM as above
Efficiency factor (called Cp) = about 0.35 for my DIY turbine
The interesting part is that wind speed is cubed, which means even a small increase in wind speed makes power rise dramatically.
Testing was the most exciting part. To find stronger wind, I climbed up to my rooftop. My little brother helped hand me the turbine and the stand while I carried my backpack full of tools and my laptop up the ladder. The roof had solar panels, water containers, and air conditioner units, so space was tight. But because of the height and the open sides with no walls blocking the wind, it was the ideal spot to set up the turbine.
At first, my ESP32 couldn’t upload data properly to the cloud, so I had to troubleshoot right on the spot. I had set it to send updates every second, but I later realized the online platform only allowed data every 15 seconds. Because of that mismatch, nothing worked at the beginning. After some testing, I changed the interval to 16 seconds, and finally the data started showing up online. It was a small fix, but it taught me how important it is to match hardware settings with cloud requirements.
The second challenge was keeping the turbine safe on the rooftop. Strong winds during a rainstorm once knocked it over, so I had to get creative with stabilizing. I tied wires to hold it steady, placed bricks to add weight, and made sure it wouldn’t slip or fall again. For the electronics, I used plastic bags to cover them as simple waterproofing. It helped, but I still found some water leaking inside during heavy rain, which showed me the limits of quick DIY protection.
Even though the setup wasn’t perfect, the system worked. The turbine spun, the sensors collected data, and I could see everything both on the LCD screen and online. Every time I solved a problem, I felt more confident that I was actually building something real and meaningful.
This project taught me that engineering isn’t just about designing on paper — it’s about adapting to real-life challenges. From 3D printing to coding, from climbing ladders to problem-solving in the rain, I learned how persistence and creativity can turn an idea into something that works. It also made me more excited about renewable energy and how technology can help create sustainable solutions for the future.