Louijie Compo
Menu

AeroLoop · Engineering note

AeroLoop finds the edge of its sensor-outage envelope

500 ms profiles recover in the retained baseline study. One-second cases are mixed and all two-second missions fail. Later prediction and landing-guard studies still report accepted: false.

Source: https://louijiecompo.com/updates/aeroloop-outage-envelope/

Sensor outage study · accepted: false
250 ms
Pass
3/3 missions
3/3 paired checks
6/6 recovery windows
500 ms
Pass
3/3 missions
3/3 paired checks
6/6 recovery windows
1 s
Mixed
2/3 missions
0/3 paired checks
3/6 recovery windows
2 s
Fail
0/3 missions
0/3 paired checks
0/6 recovery windows

Held-feedback baseline, seeds 0/1/2. Mixed means some mission passes; all one-second paired checks fail. This is not a safe-duration limit.

Download retained JSON · Source and criteria ↗

PR #19 retains fifteen measured flights under fixed mission and recovery gates. The four-column comparison below comes directly from its public JSON. It separates mission passes from paired checks and sustained recovery; the complete study reports accepted: false.

PR #20 adds opt-in prediction without closing the long-outage boundary. The latest experiment, PR #21, compares a capture-aware descent guard with prediction alone. Its 24 fresh flights add separate regression and unseen cohorts. Guarded missions pass 9/12 regression and 2/6 unseen cases, with no improvement in pass counts. Every two-second mission still fails.

The guard remains opt-in. Improved touchdown speed in some cases does not erase tracking failures or missed deadlines. Follow the source below for the latest landing experiment; the article explains the original outage envelope and both follow-ups. This portfolio review did not execute the simulator.

Follow the evidence

Explore AeroLoop

Read source evidence ↗ (opens in a new tab)

Related reading