Automated Street Light Adjustment System on Campus with AI-Assisted Data Analytics

Area of Science:

• Smart City Technology
• Artificial Intelligence
• Energy Efficiency

Background:

• Urbanization drives the need for intelligent systems in metropolitan areas.
• Optimizing energy consumption in public infrastructure is crucial.
• Smart street lighting offers a solution for efficient illumination.

Purpose of the Study:

• To develop an automatic light adjustment system for smart street lighting.
• To optimize energy efficiency while ensuring adequate campus illumination.
• To integrate AI for predictive environmental analysis and light control.

Main Methods:

• Development of device control with user interface and API.
• Creation of a prediction model using an AI-assisted data analytic platform.
• Application of four machine-learning models on a nine-month environmental dataset.

Main Results:

• The XGBoost model with a three-day window size achieved the best performance.
• A correlation coefficient of 0.922 demonstrated a strong linear relationship between actual and predicted illuminance.
• An automated system was established, operating at 20-min intervals for real-time adjustments.

Conclusions:

• The developed system effectively optimizes energy efficiency in smart street lighting.
• AI-driven prediction models enhance the accuracy of light adjustment.
• The integrated system provides a scalable solution for smart city infrastructure.