- 32 analyses: each label kept, rewritten or removed by its analyst (section Requalification of the report; mm magnitudes where the targets have them) - SNCF01 x3, WILDSTRUBEL 2026-09-01 -> 2026-10-06 x3: verified from their existing labels under the new rules - long labels (no analysis): auto/analyses/long-labels/<STATION>/; NORDEND-169 and -195 run to 07/11, so NORDEND-283 and -284 (their continuation) are removed - reading-rules.md, stations.md: the citations of removed or rewritten labels point to their successor, or to [expert] Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
3 lines
4.6 KiB
JSON
3 lines
4.6 KiB
JSON
{"id": "NEBRASKA-2", "start": "2026-10-09T08:00Z", "end": "2026-11-01T00:00Z", "family": "Target issue", "description": "The reference REFCOLH22 is lost at the 9 October noon measurement and does not come back until the end of the month: its error_value jumps almost to the maximum and all its measurements are rejected. On the thumbnails, its spot, sharp and round until the morning of 9 October, has disappeared from its place at the next measurement, while the neighboring spot and the other references are intact; what made it disappear (prism fallen, removed, hidden) is not visible on the whole image. At the same measurement, MICHIGAN, in the same building as NEBRASKA according to the expert, also loses one of its references (REFPOU22), whose spot disappears in the same way with no visible obstacle: a cause common to both, in the building, is to be looked for; it is not established. At the loss, all the targets jump together in one step: corr_x upward, most on COLJ20, POUJ19, POUJ20 and POUH19, and corr_y upward on COLH19, COLH20, COLJ23 and COLJ24; the jump does not return, it is not a movement. The compensation then runs on the two only references REFH22 and REFP''21 (the corrections of these two references cancel): a more fragile basis, which no longer compensates the camera movement well and makes all the targets move at the slightest deviation of a reference. Very bad sign for the measurement: check the prism of REFCOLH22 on site and restore it.", "review": "2026-11-17"}
|
|
{"id": "NEBRASKA-6", "start": "2026-10-16T20:00Z", "end": "2026-11-01T00:00Z", "family": "Displacement", "description": "Long upward drift of the corr_x of all the targets, strongest on the targets at the top and the right of the image (COLJ20, POUJ20, POUJ19, COLJ19, then COLH19, POUH19, COLH20) and weak in the image on those close to the references (COLH21, COLJ23, COLJ24), although COLJ24, far from the camera, drifts as much as the others in mm; the corr_y of COLH19, POUH19, POUJ20, COLJ19 and COLH20 goes down, that of POUH20 and COLH21 rises; no return at the end of the month. The camera keeps nodding forward as in the first half of the month (tilt rising, pan and rotation turning, scale dropping), and the compensation on two references only renders it poorly, so that the targets drift with it. The camera housing temperature drops from 16 October and the drift follows its jolts: fast rise until 19 October, slower until 24 October, stop and slight backward movement when the temperature rises again from 25 to 28 October, new rise with the cold of 28 and 29 October, slight backward movement when it rises again on 31 October. But the camera was already nodding forward from 3 to 12 October when the temperature was hardly dropping, and it stays more tilted at the end of the month than in the middle at equal temperature: temperature does not explain it alone, link not established. According to the expert, BREITHORN, MICHIGAN and NEBRASKA are in the same building: the cameras of the three stations tilt together, in the same direction and on the same dates (advance until 12 October, stop and slight backward movement from 13 to 16, resumption from the 17th to the end of the month), which points to a cause common to the building rather than a fault of this camera alone. Checked target by target (the raw positions against the overall movement of the targets, and a camera transformation fitted on the other targets): the scene does not move quite as a block, it deforms while the camera tilts. POUH20 moves away the most from the other targets, its neighbor POUH21 hardly at all; then COLH21, and the targets at the edges of the image (COLJ19, COLJ20, COLH19); hence the rise of the corr_y of POUH20 and COLH21. These deviations follow the camera tilt in both directions: POUH20 rises with it on 11 and 12 October, goes back down when it straightens from 13 to 16 October, rises fast with it from 17 to 19 October, then slowly until the end of the month; COLH21 only moves away from 18 October, when the camera exceeds its tilt of the 12th. It is the same movement as that of the camera, not a separate event; at equal tilt, POUH20 stays slightly higher after 17 October than before. Cause not established (the camera moving slightly as it tilts, targets at different distances, or a movement of the structure with the building). Deduced from the curves, the raw positions, the pose and the temperatures; images with nothing clear. For the customer: the camera tilts and the scene deforms with it: a movement of the structure is not established; to be monitored, the camera tilt and the curves of POUH20 and COLH21, which move away the most; restoring a third reference will make the curves reliable again."}
|