A Non-Intrusive Approach for Indoor Occupancy Detection in Smart Environments

Area of Science:

• Computer Science
• Artificial Intelligence
• Ubiquitous Computing

Background:

• Smart Environments aim to enhance user comfort by adapting conditions based on occupant presence.
• Current methods often involve intrusive data collection, necessitating non-intrusive alternatives.
• Accurate occupancy detection is crucial for effective environment adaptation.

Purpose of the Study:

• To propose and validate a non-intrusive solution for detecting and estimating the number of occupants in indoor smart environments.
• To leverage Machine Learning techniques for improved user experience in smart spaces.
• To develop a prototype system for real-world data acquisition and analysis.

Main Methods:

• Implementation of Machine Learning algorithms for occupant detection and counting.
• Development of a prototype system with integrated nodes and sensors for environmental data gathering.
• Non-intrusive data collection from indoor areas.
• Application of pattern recognition mechanisms for data analysis.

Main Results:

• The developed prototype system successfully gathered, processed, and stored environmental data.
• Machine Learning analysis of the collected data enabled accurate determination of indoor environment occupancy.
• The non-intrusive approach proved effective in identifying and estimating occupant numbers.

Conclusions:

• The proposed system effectively determines indoor occupancy using non-intrusive methods and Machine Learning.
• This approach enhances user experience in smart environments by enabling adaptive conditions.
• The research demonstrates the feasibility of real-time occupancy estimation for smart environment applications.