This is an earlier experiment. Read the amendments below before treating its findings as current.
The question
The stronger pull keeps the craft near its route, but slowly. Longer path or slower flight, and what brakes it?
What was tested
A breakdown of each steering force on re-flown runs, then the craft's speed-matching switched off under the stronger pull (two layouts).
What was found then
At least 95 % of the lost time was slower flight (21–22 m/s instead of about 48); the detours added only 2–4 % to the path. After a mislabelled column was fixed, the brake turned out to be matching the speed of slower traffic, not the pull. Switching speed-matching off doubled its speed, but it then sat about 340 m off its line and the others lost 1.7–2.5 times as many craft per craft-hour: a warning from two runs. Read then: far out, its own dodging pushed away about twice as hard as the capped pull.
What we learned later
Beyond 200 m from the route the net push points slightly toward it, and closer in slightly away: the craft settles at a balance point rather than its dodging winning a tug-of-war. One switch, one test case.
research/design/prio_pullcap_line_close.md
What changed
nothing (counters added, off)
Numbers and source references
- 8
- re-flown runsresearch/design/prio_pullcap_slow_record.md
- 0.956 / 0.971
- share of extra time due to slower speed, far cellsresearch/design/prio_pullcap_slow_record.md
- 2–4 %
- extra path from the excursionsresearch/design/prio_pullcap_slow_record.md
- 21–22 m/s vs ~48
- craft speed with vs without own-speed pull-backresearch/design/prio_pullcap_line_close.md
- 21.1 → 41.7, 21.9 → 41.9 m/s
- speed with flocking switched off, two seedsresearch/design/prio_align_off.md
- 340 m
- offset from line with flocking offresearch/design/prio_pullcap_line_close.md
- ×1.7 / ×2.5
- ordinary losses per craft-hour with its flocking offresearch/design/prio_align_off.md