This is an earlier experiment. Read the amendments below before treating its findings as current.
The question
The design's main guess was that dense traffic plus late information would make disturbances grow from craft to craft.
What was tested
The shoved hovering grid ran at three spacings between craft and four link delays, from 0 to 2 s.
What was found then
The share of the shove passed from craft to craft never rose above 0.94, so disturbances never grew (above 1 they would). Instead, close calls multiplied where tight spacing met delay. At the tightest spacing there were 283 with fresh information and 6,106 with a 2 s delay. Those counts come from the last layout only; the passed-on share is a three-layout average. The danger is craft running out of room, not a growing wave.
What changed
Nothing.
Numbers and source references
- three spacings
- 0.6, 0.75, 0.9 of the repulsion reachresults/results_phase4.json e9_phase_plane
- 0 to 2 s
- link delaysresults/results_phase4.json e9_phase_plane
- 0.94
- highest α on the plane (0.935)results/results_phase4.json e9_phase_plane
- 283
- breaches at the tightest spacing with no delayresults/results_phase4.json e9_phase_plane
- 6,106
- breaches at the tightest spacing with 2 s delayresults/results_phase4.json e9_phase_plane
- 2 s
- longest delayresults/results_phase4.json e9_phase_plane