Occupancy Estimation in Academic Laboratory: A CO2-Based Algorithm Incorporating Temporal Features for 1–16 Occupants DOI Open Access
Eliasz Kańtoch, Piotr Augustyniak

Electronics, Journal Year: 2025, Volume and Issue: 14(7), P. 1377 - 1377

Published: March 29, 2025

Private, non-intrusive presence detection methods contribute to various applications, from occupancy monitoring energy optimization and security. This study presents a deep learning approach for predicting patterns using CO2 sensor data temporal features, derived year-long dataset (18 September 2023–21 November 2024) collected via the Smart Indoor Air Quality Monitor. We created of 19,189 samples levels (0–5000 ppm) with timestamps. A sequential neural network three fully connected layers was implemented in TensorFlow. The developed model demonstrated feasibility based on features an accuracy 0.97 F1-score 0.92. Model visualization performed heatmaps. Its advantages include low computational requirements, cost-effective sensors, IoT-enabled interface, scalability. However, is limited university laboratory capacity 1–16 occupants, which may impact its generalizability other settings. These findings highlight utility estimation conditions unique, long-term multimodal research community.

Language: Английский

Multimodal feature fusion and ensemble learning for non-intrusive occupancy monitoring using smart meters DOI Creative Commons
Sakib Mahmud, Fayçal Bensaali, Muhammad E. H. Chowdhury

et al.

Building and Environment, Journal Year: 2025, Volume and Issue: unknown, P. 112635 - 112635

Published: Jan. 1, 2025

Language: Английский

Citations

0
Occupancy Estimation in Academic Laboratory: A CO2-Based Algorithm Incorporating Temporal Features for 1–16 Occupants DOI Open Access
Eliasz Kańtoch, Piotr Augustyniak

Electronics, Journal Year: 2025, Volume and Issue: 14(7), P. 1377 - 1377

Published: March 29, 2025

Private, non-intrusive presence detection methods contribute to various applications, from occupancy monitoring energy optimization and security. This study presents a deep learning approach for predicting patterns using CO2 sensor data temporal features, derived year-long dataset (18 September 2023–21 November 2024) collected via the Smart Indoor Air Quality Monitor. We created of 19,189 samples levels (0–5000 ppm) with timestamps. A sequential neural network three fully connected layers was implemented in TensorFlow. The developed model demonstrated feasibility based on features an accuracy 0.97 F1-score 0.92. Model visualization performed heatmaps. Its advantages include low computational requirements, cost-effective sensors, IoT-enabled interface, scalability. However, is limited university laboratory capacity 1–16 occupants, which may impact its generalizability other settings. These findings highlight utility estimation conditions unique, long-term multimodal research community.

Language: Английский

Citations

0