SuryaNetra

Inspiration

Solar panels lose a significant amount of potential energy due to improper orientation, dust accumulation, bird droppings, and changing environmental conditions. Most existing tracking systems rely only on basic sensors and do not account for surface cleanliness or real-world inefficiencies. We wanted to build a smarter system that not only follows the sun but also understands the physical condition of the panel and reacts intelligently. SuryaNetra was inspired by the idea of giving solar panels vision and intelligence so they can actively maximize their own efficiency.

What it does

SuryaNetra is an intelligent solar tracking and monitoring system that automatically aligns a solar panel toward the sun, detects dust and bird droppings on the panel surface using AI-based vision, and triggers cleaning when required. The system continuously monitors voltage, current, and power output, sends real-time data to the cloud, and displays performance metrics on a web dashboard. It improves energy generation by combining sensor-based tracking, computer vision, and intelligent decision-making.

How we built it

We used an STM32 or ESP32 microcontroller to read data from light sensors and power sensors through ADC channels. Motors are controlled using PWM to rotate the panel along horizontal and vertical axes. A camera module captures images of the panel surface. Images are processed using OpenCV, and a lightweight CNN model is used to classify dust and bird droppings. The microcontroller communicates with the cloud using Wi-Fi and sends telemetry data through MQTT or HTTP. A web dashboard built with JavaScript displays real-time graphs, system status, and alerts.

Challenges we ran into

Sensor readings were noisy due to varying sunlight conditions. Motor overshoot caused oscillations during tracking. Running AI models on limited hardware was difficult. Wi-Fi connectivity was unstable at times. We solved these by applying filtering techniques, introducing dead zones in motor movement, using lightweight models, and adding communication retries.

Accomplishments that we're proud of

We successfully integrated solar tracking, AI-based surface condition detection, cloud connectivity, and dashboard visualization into one working system. We achieved real-time tracking with stable motor control and reliable data transmission. The system demonstrates measurable improvement in energy output and autonomous maintenance capability.

What we learned

We gained experience in embedded systems programming, sensor interfacing, motor control, computer vision, machine learning integration, and cloud-based IoT systems. We also learned how to design hardware and software together, debug real-time systems, and optimize solutions under resource constraints.

What's next for SuryaNetra

We plan to add weather-aware prediction, improve model accuracy with larger datasets, integrate a fully automated cleaning mechanism, develop a mobile application, and support large-scale multi-panel installations.

Built With

• amazon-web-services
• embedded
• mqtt
• rest