Calibrated sky, not footage
Bench to university. · 2 min read
What it proposes
Every argument about anomalous objects in the sky is really an argument about instrumentation. A passive electro-optical or infrared sensor cannot supply range, so a size and a speed derived from one camera depend entirely on an assumed distance. The fix is not better arguing, it is two of every sensor, declared thresholds, and control sites — and three groups have now published exactly that standard. The programme worth running is the one that turns kinematic claims into calibrated measurements with error bars.
Who it is forInstrumentation engineersOptical and infrared systems designersData-pipeline builders
Why the library suggests it
Kevin Knuth, Robert Powell and Peter Reali take well-documented encounters and compute the smallest acceleration each observed manoeuvre could possibly have involved, with generous uncertainties; the lower bounds run from about 68 g to well over 5000 g, against about 13.5 g for a fighter airframe and 45 g for a human being for a twentieth of a second — and the striking part is what is missing alongside them, no air disturbance, no sonic boom, and no heat consistent with the implied power (Estimating flight characteristics of anomalous unidentified aerial vehicles, 2019). A field expedition then reported exactly what its equipment recorded and exactly what it could not settle — about an hour of triggered visible and night-vision video, more than 600 hours of continuous far-infrared, 55 hours of radiation data, one unidentified event, ten ordinary explanations listed, and the observation that a second camera would have eliminated half of them — and turned the experience into a usable standard: two of every sensor, three sigma for an ambiguity, five for an anomaly (Initial results from the first field expedition of UAPx, 2025). The observatory-scale version of the same discipline, with categories declared before the data arrives and everything routed through peer review, is Overview of the Galileo Project (2023).
The experiment or build
Build a two-station observing node: paired visible and infrared cameras at each station, separated by a known baseline, time-synchronised, with a magnetometer and a particle counter on the same clock. Stereo gives range, and range is what turns an angular measurement into a size and a speed. The settling measurement is a triangulated range, and therefore a true size and acceleration with error bars, for any object recorded at both stations — published together with the detection threshold and the number of observing hours, so the null hours count as data. Reporting the hours is what makes the rate meaningful.
Where it stands
On the bench now — instrumented expeditions are running, the standards are published, and the stereo-baseline network they call for has not been built at scale.
Take it up
- The measurement that settles it
- The settling measurement is a triangulated range, and therefore a true size and acceleration with error bars, for any object recorded at both stations — published together with the detection threshold and the number of observing hours, so the null hours count as data.
- What it costs to start
- Bench to university.
- The engineer it grows
- It is also the card that teaches the most transferable habit in this programme: state your threshold before you look.
What it rests on
Where it sits in the curriculum
The unified pictureInertial mass reduction, the Navy patents and transmedium craftThe reference documents and the institutional record