This is an earlier experiment. Read the amendments below before treating its findings as current.
The question
Keep full extra room for pairs drifting together slowly (crowded, gusty airspace) and shrink it for pairs closing fast (crossings), keeping both gains. That needs the two kinds of pair to close at different speeds.
What was tested
After a reviewed specification, 73 existing runs (wing crossing, multirotor storm test) were replayed. Closing speed was recorded at the start of each serious approach (over halfway into the spacing), as it really was and as the controller heard it over the radio link.
What was found then
The yardstick was how much room each approach would lose under the proposal. In reality the two situations were about the same (0.53 crossing, 0.57 crowded airspace), the opposite of the predicted gap. As the controller heard it, the crossing scored slightly higher, but by less than the 0.2 gap needed. The gap was about 0.4 at 8 m/s but only 0.07 at 22 m/s, where the wing's losses are. On this limited sample the idea had nothing to act on; it was shelved unbuilt.
What changed
nothing (a closing-speed log, off by default, was added to the controller for measurement only)
Numbers and source references
- 73
- runs measured, each reproducing its earlier collision countresearch/data/oos/tau_closure_census.md
- 0.53 vs 0.57
- average physical reduction share (credit), crossing vs gusty airspace, reference thresholdresearch/data/oos/tau_closure_census.md
- 0.16 / 0.11 / 0.06
- crossing minus airspace credit in the controller's view at three thresholdsresearch/data/oos/tau_c0_census.md
- 0.2
- gap the idea neededresearch/data/oos/tau_c0_census.md
- about 0.4; 0.07
- controller-view gap at 8 m/s and 22 m/s (0.85−0.46; 0.53−0.46)research/data/oos/tau_c0_census.md