7 hours of strawberries: the humidity surprise

The hypothesis

Most spoilage sensors only watch gas. But microbes don’t only release gases. They also release water, because their metabolism is wet chemistry. If I sealed strawberries in a jar, would I see humidity climb even before the gas-resistance curve dropped?

The setup

• 6 strawberries, dry, no visible mold
• Glass mason jar, sealed
• BME688 mounted on the lid (it measures both gas resistance AND relative humidity)
• 7-hour test, room temperature, kitchen counter

What happened

Hour	RH %	gas kΩ
0	58.4	24.8
1	58.9	24.7
2	59.3	24.6
3	59.8	24.5
4	60.2	24.4
5	60.6	24.2
6	60.9	24.0
7	61.2	23.7

Gas resistance moved 1.1 kΩ, about a 4% drop. Humidity moved 2.75 percentage points, a much larger relative change. The humidity signal was easier to detect than the gas signal in the first few hours.

Why this matters

For some foods (especially produce), humidity might be a better early-warning signal than gas. The microbes have to grow up to a population large enough to produce detectable gas concentrations, but they’re releasing water vapor from the very first cell divisions.

V6 firmware will treat humidity-rise-rate as its own input channel, not just a sensor calibration correction. If the BME688 sees humidity climbing faster than 0.3 %/hour in a sealed container, the system will flag “early spoilage activity” even if the gas reading is still nominal.

What I didn’t expect

The control jar (an empty sealed jar of the same size) showed humidity creeping up by 0.4 percentage points over the same 7 hours, just from the residual moisture in the room. So I have to subtract the baseline drift. The strawberry signal is real but it’s smaller than the raw number suggests. Real science: always check your control.

• Sri