How the Weights Decide the “Best” Sensor
How the Weights Decide the “Best” Sensor
Ada re-derives this chapter’s own numbers step by step, at full precision
ADA · CALCULATION AUDIT
How the Weights Decide the "Best" Sensor
The chapter scores IoT temperature sensors with weighted requirements — accuracy at 9, waterproofing at 10, sleep current at 7, cost at 5, and humidity at 3, for a total possible weight of 34 — and reports the DS18B20 at 91% against the DHT22’s 50%, a 41-point gap. Those percentages are presented as the deciding argument for “best” sensor. This audit asks the question the gap invites: does the arithmetic behind 91% and 50% actually check out, and where does the 41-point gap really come from?
Companion to the chapter Choosing the Right Sensor — every number here comes from that chapter.
See the relationship before changing it
The figure reads from left to right. The blue card is waterproofing weight. The middle card applies this page's rule. The green card is ds18b20 score. Walk the arrows once: set the input, apply the rule, then read the result with its unit.
The retained audit below checks several chapter fixtures. This model keeps those stated values fixed and changes only waterproofing weight, so the numeric fixture does not switch without explanation.
Derive the baseline in four named moves
- 1
Name the input. The chapter baseline is 10 points.
- 2
Name the relationship. score = (21 fixed points + waterproofing weight) / (24 fixed possible points + waterproofing weight) x 100
- 3
Substitute with units. (21 + 10) / (24 + 10) x 100 = 91.2%
- 4
Read the result. Keep the unit beside the value. Use it only inside the technical boundary on this page.
Predict, then change waterproofing weight
Try Predict the direction of score = (21 fixed points + waterproofing weight) / (24 fixed possible points + waterproofing weight) x 100. Test another waterproofing weight, then compare ds18b20 score.
Observe A heavier waterproofing requirement lifts this sensor because it passes that test, but the declared weights still manufacture the ranking. Reset waterproofing weight to 10 and compare ds18b20 score.
Explain A heavier waterproofing requirement lifts this sensor because it passes that test, but the declared weights still manufacture the ranking.
Check yourself
What should you do before trusting a moved-control result?
What does this small model leave out?
Sum the requirement weights 9 + 10 + 7 + 5 + 3 = 34, then score DS18B20, DHT22, and BME280 against the same checklist.
The totals become 31/34 = 91.2%, 17/34 = 50.0%, and 15/34 = 44.1%; waterproofing plus sleep current create 17 points of the leader's gap.
A weighted matrix converts declared priorities into the ranking, so changing or removing a high-value requirement changes the winner even when no datasheet value changes.
Ada: The weighted-scoring section reports DS18B20 at 91% and DHT22 at 50%, with a 41-point gap. Those percentages are the whole argument, so let me rebuild them from the weights: accuracy 9, waterproof 10, sleep 7, cost 5, humidity 3, for a total possible of 9 + 10 + 7 + 5 + 3 = 34.
- DS18B20 passes accuracy, waterproof, sleep, cost:
9 + 10 + 7 + 5 = 31, so31 / 34 x 100 = 91.2% - DHT22 passes accuracy, cost, humidity:
9 + 5 + 3 = 17, so17 / 34 x 100 = 50.0% - BME280 passes sleep, cost, humidity:
7 + 5 + 3 = 15, so15 / 34 x 100 = 44.1%
The gap is 91.2 - 50.0 = 41.2 points, and it comes almost entirely from the two heaviest requirements DHT22 misses – waterproof (10) and sleep current (7) sum to 17 of the 34 possible points. The design-meaningful conclusion is that the ranking is manufactured by the weights, not discovered in the datasheets: set waterproofing’s weight to zero for an indoor project and DS18B20 surrenders its 10-point lead, which is exactly why the chapter insists there is no objectively best sensor, only the best sensor for a stated priority vector.
Every number above is taken from the chapter’s own material and re-derived step by step.
Technical boundaries: This binary pass/fail scorecard omits measurement uncertainty, partial compliance, correlated criteria, lifecycle and integration cost, supplier risk, calibration drift, and sensitivity analysis across alternative weights.