Relationship between Yahi and PSI

Yahi Overview
Relationship between Yahi and PSI
Mapping Pilot
Assembly Instructions
Installation Instructions
Developers
Data
FAQ

Although Yahi and PSI are both concerned with air quality, they have different objectives, use different approaches and therefore have different results.

	PSI Measurement	Yahi Project
Objective	human health impact of airborne pollutants	mapping the location and movement of haze
Accuracy	high (lab-grade instruments, operated by experts, years of research)	limited (consumer-grade dust sensors, unsupervised operation, very limited dataset for current PM2.5 estimation)
Measurement Period	24 hours (sometimes shorter)	30 seconds (noisy data with very high spikes)
Pollutant Scope	particles, SO₂, CO, O₃, NO₂	particles only
PM2.5 Assessment Approach	collect particles smaller than 2.5 microns for 24 hours weigh them measure PM2.5 accurately	count approx 1 micron particles for 30 seconds (particles smaller than 1 micron are not detected) estimate PM2.5 based upon historical correlation between Yahi counts and NEA PM2.5 measurements

From time to time I hear people describe Yahi’s approach as being more accurate than that performed with lab-grade instruments. Hopefully the above makes clear not only that this is not the case, but also that it cannot possibly be the case. “Big data” analytics make all sorts of things possible - and enough data will eventually remove one of the qualifiers above - but the rest will remain, so Yahi’s PM2.5 estimates will always be estimates only.

I also occasionally hear people - often the same people - describe the Yahi project as being independent of the NEA. This is only half true. The sensors and data collection/processing/display are independent of the NEA, however the historical PM2.5 data that’s used for correlation analysis and PM2.5 estimation is clearly not. Yahi is possible in large part because of the NEA’s publication of accurate data about air pollutants in Singapore.