EN
Back to the archive

The encyclopedia · Product & Design · Product decision · 2024

EarSwitch's oximeter reads oxygen in the ear to remove skin-tone bias.

EarSwitch's Airquity campaign shows finger pulse oximeters misread darker skin and offers an inner-ear oximeter that does not.

EarSwitch · Havas Lynx

the move

Finger pulse oximeters are meant to save lives, but they are less accurate for people with brown and black skin, and a few percent of error can be the difference between a normal reading and a life-threatening alert.

EarSwitch, with agency Havas Lynx, built Airquity to make that bias visible. The campaign pointed out that finger-clip oximeters overestimate oxygen in darker skin and that the problem has been documented since at least the 1990s.

The product behind it is EarMetrics-Oximetry, which measures oxygen in the inner ear canal, where skin colour is consistent across people. That removes the variable that corrupts the reading rather than patching it.

The Airquity campaign took two Silver Lions at Cannes Lions 2024, using the platform itself to get the bias on the record and position the ear-based device as the equitable fix.

why it works

  • The ear canal's skin is independent of external skin tone, so the confound disappears
  • A few percent of error can flip a life-or-death decision, so the bias is not cosmetic
  • It addresses the root cause of the inaccuracy rather than the calibration
  • The campaign amplifies the message beyond the brand's own marketing budget
the payoffRead the signal where skin tone does not varyclever

what transfers

When a sensor systematically errs on one group, move the measurement to a place where the confounding variable disappears rather than trying to recalibrate it away.

what came after

Airquity won two Silver Lions at Cannes Lions 2024, putting the racial bias in finger oximetry in front of a global audience and positioning EarMetrics-Oximetry as a potential answer. The device aimed to make oxygen readings racially equitable, though adoption beyond the campaign was still to be proven.

references

spotted an error? The archive wants to know.

same kind of clever