Dark Mode
Environmental Metrics for Occupant Count Estimation
Environmental Monitoring
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
Estimating the number of occupants within an indoor environment using non-intrusive environmental markers provides a privacy-preserving alternative to camera-based surveillance. This collection records data from a series of sensor nodes deployed in a 6m x 4.6m room, capturing variations in temperature, light, sound, CO2 levels, and motion. By transmitting data every 30 seconds to a central edge node, the setup monitored occupancy levels ranging from zero to three people. This information is vital for developing smart building systems that can optimise energy usage and space management based on real-time human presence, especially as the data was collected in a controlled setting without active HVAC systems.
Columns
- Date: The calendar date of the recording, spanning from December 2017 to January 2018.
- Time: The specific time of day for each recorded 30-second interval.
- S1_Temp / S2_Temp / S3_Temp / S4_Temp: Temperature readings from four different sensor nodes measured in degrees Celsius.
- S1_Light / S2_Light / S3_Light / S4_Light: Ambient light levels recorded by four separate nodes within the room.
- S1_Sound / S2_Sound / S3_Sound / S4_Sound: Analog sound levels captured as voltage readings, adjusted via amplifiers for peak sensitivity.
- S5_CO2: Carbon dioxide concentration levels measured by a dedicated sensor node that underwent manual zero-point calibration.
- S6_PIR / S7_PIR: Digital passive infrared sensor readings indicating motion detection from ceiling-mounted nodes.
- Occupancy_Count: The manually recorded ground truth indicating the precise number of people in the room (0, 1, 2, or 3).
Distribution
The data is delivered in a structured CSV format titled
Occupancy_Estimation.csv, with a file size of 931.63 kB. It contains 10,100 records, each representing a discrete measurement interval. The file displays high integrity with a 100% validity rate across all recorded columns, ensuring no missing or mismatched data points. This resource holds a usability score of 10.00 and is provided as a static archive with no future updates planned.Usage
This resource is ideal for training and validating machine learning models designed for occupancy estimation and multi-class classification. It is well-suited for research into sensor fusion, where data from diverse sources like CO2 and light sensors are integrated to improve prediction accuracy. Additionally, developers can use the records to test smart building automation logic, such as adjusting lighting or ventilation based on detected occupant counts.
Coverage
The geographic scope is limited to a single experimental testbed located in a 6m x 4.6m room. Temporally, the records were collected over a four-day period between 22 December 2017 and 11 January 2018. The demographic coverage includes controlled occupancy variations between zero and three people. Notably, the experiment was conducted without HVAC systems to prevent external airflow from influencing the environmental sensor readings.
License
Attribution 4.0 International (CC BY 4.0)
Who Can Use It
IoT engineers can leverage these records to refine the calibration and sensitivity of environmental sensor networks for indoor tracking. Data scientists may utilise the clean, high-integrity records to practice feature engineering and classification algorithms on real-world sensor data. Furthermore, academic researchers in the fields of smart cities and green building design can use this data to model the relationship between human presence and environmental changes.
Dataset Name Suggestions
- Room Occupancy Estimation: Multi-Sensor Environmental Registry
- Non-Intrusive Indoor Occupancy and Environmental Sensor Archive
- Smart Building Occupancy Tracking: 7-Node Sensor Network Data
- Environmental Metrics for Occupant Count Estimation
- PIR, CO2, and Temperature Records for Room Occupancy Modelling
Attributes
Original Data Source: Environmental Metrics for Occupant Count Estimation
QUALITYLoading...
Free
Download Dataset in CSV Format
Recommended Datasets
Loading recommendations...