AI-Enabled Indoor Air Quality Monitoring for Fungi Energy

Introduction

The COVID-19 pandemic heightened awareness of indoor air quality as a public health priority. Poor IAQ is linked to impaired health, reduced productivity and growing organisational liability.

Sensus Air, operating as AIRQ108 Pty Limited, is a South African AI-driven IAQ monitoring venture. This study documents its deployment with Fungi Energy from November 2020 to June 2023.

Background

Sensus Air integrates non-intrusive sensors with machine learning for real-time environmental risk classification across corporate, healthcare and manufacturing sectors.

Fungi Energy is a progressive energy company whose offices are open-plan with centralised air conditioning. The company recognised the need to address IAQ challenges arising from high occupancy and limited ventilation, the combination that makes open-plan floors difficult to condition well.

Methodology

Sensors measured temperature, humidity, particulate matter (PM1, PM2.5, PM4, PM10 in µg/m³), gaseous pollutants (CO and CO₂ in ppm, VOCs including benzene and formaldehyde), and spectral signatures analysed by machine learning.

The pilot was a structured assessment of corporate office IAQ under centralised cooling over a 2.5-year period, measuring baseline conditions and mitigation potential.

• Data collection from deployed sensors
• Pattern recognition using machine learning
• Real-time dashboard reporting
• Targeted mitigation recommendations

Results: carbon dioxide

Average concentration was 2,500 ppm, with a secondary reading of 2,030 ppm. Although classified as normal by the index in use, this approaches levels indicating inadequate ventilation and sits well above the 1,000–1,500 ppm range associated with optimal cognitive performance.

For an open-plan floor on centralised cooling, this is the signature of air being recirculated rather than replaced.

Results: particulate matter

• PM1, high peaks indicating fine particle accumulation
• PM2.5, elevated levels posing respiratory health risks
• PM4 and PM10, normal levels, suggesting effective larger particle control

Results: hazardous volatile organic compounds

Spectral analysis named the compounds present and graded them by concentration.

Environmental KPIs

• Pathogen Risk Index, normal for CO₂, PM4 and PM10
• Pathogen Risk Index, elevated for PM2.5 and VOCs

Discussion

Elevated PM2.5 and VOC levels necessitate immediate mitigation, and the presence of carcinogenic compounds underscores the urgency. Spectral signature analysis provides an advantage in real-time pollutant identification and source contextualisation, enabling targeted health and productivity improvements alongside ESG compliance.

Limitations

• Single open-plan office environment, limiting facility-level representation
• No quantification of the effectiveness of specific mitigation measures
• Seasonal variation and occupancy pattern influences not addressed
• Longitudinal analysis needed

Conclusions and recommendations

Elevated PM2.5 and VOC levels were identified alongside carcinogenic compound risks. The recommendations below follow directly from the source attribution the spectral analysis made possible.

• Expand sensor deployment across Fungi Energy facilities
• Implement air purifiers, enhanced ventilation and low-VOC materials
• Integrate real-time dashboards
• Run follow-up longitudinal studies evaluating mitigation effectiveness