Mapping Occupational Hazards with a Multi-sensor Network in a Heavy-Vehicle Manufacturing Facility

Area of Science:

• Environmental Science
• Occupational Health
• Sensor Technology

Background:

• Low-cost sensors enable spatially distributed sensor networks for environmental monitoring.
• Existing networks are often deployed in ambient environments.
• Industrial settings present unique challenges for continuous hazard monitoring.

Purpose of the Study:

• To develop and deploy a multi-hazard sensor network in a heavy-vehicle manufacturing facility.
• To assess the temporal and spatial distribution of key occupational hazards.
• To evaluate the precision and accuracy of low-cost sensors in a real-world industrial environment.

Main Methods:

• A 40-node wireless sensor network was constructed using low-cost sensors for particulate matter, carbon monoxide, oxidizing gases, and noise.
• Data were collected at 5-min intervals over an 8-month continuous deployment period.
• Network measurements were compared against field reference instruments to determine accuracy.

Main Results:

• The network successfully recorded temporal and spatial hazard data, revealing diurnal, weekly, and daily patterns.
• Mean hazard levels during production were: 0.62 ± 0.2 mg m-3 (PM), 7 ± 2 ppm (CO), 155 ± 58 ppb (OX), and 82 ± 1 dBA (noise).
• Sensor precision varied, with 95% of measurements within 0.21 mg m-3 (PM), 0.4 ppm (CO), 9 ppb (OX), and 1 dBA (noise) of collocated nodes; median biases were 27% (PM), 11% (CO), 45% (OX), and 1% (noise).

Conclusions:

• The study demonstrates the feasibility of long-term multi-hazard sensor network deployment in industrial settings.
• High temporal and spatial resolution hazard data can be obtained, offering comprehensive occupational exposure assessments.
• This approach paves the way for individual-level occupational exposure characterization using wireless sensor networks.