AirGap

Quantitative evidence

Why our method is better

Four pieces of evidence in order, strongest first. All of them draw on the same candidate pool — the 50,813 areas that pass the installation eligibility filter — so the rival strategies were never handed worse options.

Evidence 1

Adding sensors at our recommended sites genuinely improves the model

+2.1%model error falls by this much once our sensors are added

Only our strategy consistently reduces model error. But with 8 repetitions, the direction of the effect is clear while its magnitude cannot yet be claimed with confidence.

StrategyError reductionRuns improvedError after
Our method+2.1% ± 5.45/820.75
Most populated cities+0.7% ± 6.44/821.01
Random placement-0.7% ± 7.84/821.29

Result of each run

One dot per repetition. The spread is wide, and we show it precisely because of that.

Our method
+2.1%
Most populated cities
+0.7%
Random placement
-0.7%
15%0+15%

How this was tested

The 16 independent locations are shuffled into three roles: 3 are hidden to measure error, 4 become candidate sites for new sensors, and the remaining 9 form the training data. Each strategy picks 2 of the 4 candidates, the model is retrained, and error is measured again on the 3 hidden locations. The whole process repeats 8 times with different splits.

The standard deviation is still ±5.4 across only 8 repetitions, so this result is not statistically significant. The direction is consistent; the magnitude is not established. We report it as it stands.

Evidence 2

Our sensors do not overlap each other

18 of 20sensors genuinely stand alone

Nearly every one of our sensors measures a different body of air. Under the most-populated-cities strategy, only 2 of 20 stand alone — the rest re-measure air another sensor already covers.

StrategyStand aloneWastedArea typesClosest pairPopulation
Our method18/2010%1094.6 km561,912
Random placement17.6/2012.2%7.953.0 km156,059
Most populated cities2/2090%24.6 km8,659,761

The definition used

Two sensors count as overlapping when they sit closer than 104.4 km, the spatial range of influence we measured from the data ourselves. Effective clusters is the number of groups left after nearby sensors are merged into one.

Our method against 30 random runs

Each bar is one random run, sorted. The red line marks where our method lands.

Our method 94.6

27/30 random runs scored lower than our method.

One figure that does not flatter us

The city-based strategy wins decisively on population covered: 8.66 million people against our 561,912. We show it because cherry-picked numbers are not evidence.

But that is exactly the problem. 2 of its 20 sensors pile into the same metropolitan area and measure much the same thing over and over. Its population coverage looks large on paper; the new knowledge it buys is close to zero.

Evidence 3

Without spatial spreading, the result is far worse

+45%population covered actually rises when sensors are spread out

Spreading sensors out does not cost population coverage — it raises it by 45%.

Top scores onlyWith spatial damping
Area types reached38
Overlapping pairs700
Closest pair5 km333 km
Population covered412,418598,201

Had we simply taken the top 20 scores with no distance damping, they would bunch together: 70 of 190 site pairs overlapping, and area-type coverage collapsing to 3.

Evidence 4

The effect of the regime swap

6 of 8never-measured area types are now reached

We deliberately swapped the three lowest-scoring sites for candidates from area types that have never been measured. Coverage doubled at the cost of a 6% drop in population — a trade-off we made knowingly.

Before the swapAfter the swap
Area types reached810
Never-measured types36
Population covered598,201561,912
Closest pair333 km94.6 km

The minimum-distance threshold does not drive the result

We tested the minimum-distance threshold at 8 different values, from 0 to 300 km. All of them produced 20 identical sites with exactly the same total score. The distance rule works purely as a safety net, not as a parameter tuned to flatter the outcome.

0 km → 20/2025 km → 20/2052 km → 20/2075 km → 20/20104 km → 20/20150 km → 20/20200 km → 20/20300 km → 20/20