auto: the long labels requalified as analyses like the others, <STATION>_<start>_<end>/ with request.json (ex long-labels/<STATION>/); their labels carry the analysis; work/ out of git
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
1 parent
ddb17c2189
commit
764af7d8ec
132 files changed
+125
-9256
No files matched your search
@@ -1,8 +1,8 @@
|
||||
{"id": "WILDSTRUBEL-46", "start": "2026-09-07T00:00Z", "end": "2026-10-01T00:00Z", "family": "Target issue", "description": "The reference R3 is rejected on every measured day of the batch (from 07 to 09/09, then from 22 to 30/09), from sunrise until mid-afternoon, and almost never at night: by day, its error_value stays above its threshold at almost every measurement, a little less often in the last days of September; at night, it is well below. On its daytime thumbnail (30/09 at 10:00), its spot cannot be seen, while those of R1 and R2 can; at night, under the infrared, it is sharp, smaller than theirs (03/09 at 02:00 thumbnail). It is rejected as much before as after the outage from 09/09 to 22/09, when the other reflectors are better seen: it is the target, not the daytime sun. The compensation therefore runs every day on five references, and on three or fewer as soon as two others are lost (PA_8 poorly seen before the outage, passages in front of R1 and R2, sunrise, separate labels); its entries and exits do not make the targets jump. Cause not established: a reflector too weak for daytime (small, dirty or poorly oriented). Ignore the curve of R3 by day; act on the target (clean it, reorient it or replace it) to give a sixth reference back to the daytime compensation.", "review": "2026-11-17"}
|
||||
{"id": "WILDSTRUBEL-69", "start": "2026-09-09T21:45Z", "end": "2026-09-22T08:45Z", "family": "No data", "description": "No measurement for any sensor for nearly two weeks; the metrics and the images stop with the measurements. Just before, the battery empties and the voltage falls well below its level of the other nights (metrics): the station shuts down for lack of energy. It restarts on 22/09 in the morning, the battery full and a voltage much higher than usual, as on an external power supply; on return, the tilt of the camera is no longer at the same night level as before the outage. The curves are missing: the station ran out of energy, to be checked (battery, panel, possible intervention).", "review": "2026-11-17"}
|
||||
{"id": "WILDSTRUBEL-70", "start": "2026-09-09T21:45Z", "end": "2026-09-22T09:00Z", "family": "Displacement", "description": "During the outage from 09/09 to 22/09, the corrected curves change level and do not resume the one from before: corr_y of PB1_2 markedly lower, corr_y of PB2_2 lower, corr_x of PB_5, PA2_2, PA3_2 and PB4 lower, corr_y of PA_9 higher, from the first measurement. The jump is hidden in the outage; the step lasts until the end of the batch, except for PB_5, back near its level from before on 30/09 (labelled separately), and PB1_2 keeps going down from 25 to 29/09 (labelled separately). The camera is no longer at the same night level (tilt and rotation shifted). Relative to the overall movement of the targets, the raw position of PB1_2 has shifted alone, and the references PA_8 and PA1_2 a little relative to each other, which changes the basis of the other targets. After the restart, the reflectors are seen better, by day and by night, and the error_value of all sensors is lower; the housing temperature has the same cycle before and after. Cause not established: an intervention on the station during the outage (adjustment, cleaning), or, for PB1_2, a movement of the kerb. For the targets that follow the basis, this is not a movement of the structure: do not compare the levels from before and after the outage, to be checked with the operator; for PB1_2, a movement is not excluded: look at the target, keep watching.", "review": "2026-11-17"}
|
||||
{"id": "WILDSTRUBEL-130", "start": "2026-10-01T00:00Z", "end": "2026-11-01T00:00Z", "family": "Target issue", "description": "The reference R3 is rejected every measured day, from the first measurement of the morning (04:30 to 05:45) until 13:30-16:30, and almost never at night: most of the measurements from 05:00 to 15:00, often all of them, from 01/10 to 09/10, and at the only daytime measurement of 15/10 (08:00); almost none at night. Its median error_value is markedly above the threshold by day, and a little below it at night. It is as much so on the days when the housing heats up least (on 08/10, almost all its daytime measurements): it is not the daytime sun, it is the target. On its daytime thumbnails, its spot cannot be seen (30/09 at 10:00), while those of R1 and R2 can. The compensation therefore runs every day on five references, and on three or fewer when two others are lost (passages, PA_8 and PA1_2 poorly seen in the sun, separate labels); its entries and exits do not make the targets jump (median jump of the corr no more than a little above the noise, like the other steps of the day, apart from the moments already labelled). No daytime measurement after 15/10: no return seen, until the end of the batch. Cause not established: a reflector too weak for daytime (small, dirty or poorly oriented). Ignore the curve of R3 by day; act on the target (clean it, reorient it or replace it).", "review": "2026-11-17"}
|
||||
{"id": "WILDSTRUBEL-199", "start": "2026-10-15T08:00Z", "end": "2026-10-31T20:15Z", "family": "No data", "description": "No measurement for any sensor for more than two weeks; the metrics and the images are also missing almost all that time (two isolated values on 25/10), and come back only on 30/10 in the evening, the battery well charged but not full; it is full only on 31/10 in the morning, with daytime images on 31/10, still without a displacement measurement until that evening. Cause not established (station stopped). The curves are missing: to be checked with the operator."}
|
||||
{"id": "WILDSTRUBEL-46", "start": "2026-09-07T00:00Z", "end": "2026-10-01T00:00Z", "family": "Target issue", "description": "The reference R3 is rejected on every measured day of the batch (from 07 to 09/09, then from 22 to 30/09), from sunrise until mid-afternoon, and almost never at night: by day, its error_value stays above its threshold at almost every measurement, a little less often in the last days of September; at night, it is well below. On its daytime thumbnail (30/09 at 10:00), its spot cannot be seen, while those of R1 and R2 can; at night, under the infrared, it is sharp, smaller than theirs (03/09 at 02:00 thumbnail). It is rejected as much before as after the outage from 09/09 to 22/09, when the other reflectors are better seen: it is the target, not the daytime sun. The compensation therefore runs every day on five references, and on three or fewer as soon as two others are lost (PA_8 poorly seen before the outage, passages in front of R1 and R2, sunrise, separate labels); its entries and exits do not make the targets jump. Cause not established: a reflector too weak for daytime (small, dirty or poorly oriented). Ignore the curve of R3 by day; act on the target (clean it, reorient it or replace it) to give a sixth reference back to the daytime compensation.", "review": "2026-11-17", "analysis": "WILDSTRUBEL_2026-09-07T0000_2026-11-01T0000"}
|
||||
{"id": "WILDSTRUBEL-69", "start": "2026-09-09T21:45Z", "end": "2026-09-22T08:45Z", "family": "No data", "description": "No measurement for any sensor for nearly two weeks; the metrics and the images stop with the measurements. Just before, the battery empties and the voltage falls well below its level of the other nights (metrics): the station shuts down for lack of energy. It restarts on 22/09 in the morning, the battery full and a voltage much higher than usual, as on an external power supply; on return, the tilt of the camera is no longer at the same night level as before the outage. The curves are missing: the station ran out of energy, to be checked (battery, panel, possible intervention).", "review": "2026-11-17", "analysis": "WILDSTRUBEL_2026-09-07T0000_2026-11-01T0000"}
|
||||
{"id": "WILDSTRUBEL-70", "start": "2026-09-09T21:45Z", "end": "2026-09-22T09:00Z", "family": "Displacement", "description": "During the outage from 09/09 to 22/09, the corrected curves change level and do not resume the one from before: corr_y of PB1_2 markedly lower, corr_y of PB2_2 lower, corr_x of PB_5, PA2_2, PA3_2 and PB4 lower, corr_y of PA_9 higher, from the first measurement. The jump is hidden in the outage; the step lasts until the end of the batch, except for PB_5, back near its level from before on 30/09 (labelled separately), and PB1_2 keeps going down from 25 to 29/09 (labelled separately). The camera is no longer at the same night level (tilt and rotation shifted). Relative to the overall movement of the targets, the raw position of PB1_2 has shifted alone, and the references PA_8 and PA1_2 a little relative to each other, which changes the basis of the other targets. After the restart, the reflectors are seen better, by day and by night, and the error_value of all sensors is lower; the housing temperature has the same cycle before and after. Cause not established: an intervention on the station during the outage (adjustment, cleaning), or, for PB1_2, a movement of the kerb. For the targets that follow the basis, this is not a movement of the structure: do not compare the levels from before and after the outage, to be checked with the operator; for PB1_2, a movement is not excluded: look at the target, keep watching.", "review": "2026-11-17", "analysis": "WILDSTRUBEL_2026-09-07T0000_2026-11-01T0000"}
|
||||
{"id": "WILDSTRUBEL-130", "start": "2026-10-01T00:00Z", "end": "2026-11-01T00:00Z", "family": "Target issue", "description": "The reference R3 is rejected every measured day, from the first measurement of the morning (04:30 to 05:45) until 13:30-16:30, and almost never at night: most of the measurements from 05:00 to 15:00, often all of them, from 01/10 to 09/10, and at the only daytime measurement of 15/10 (08:00); almost none at night. Its median error_value is markedly above the threshold by day, and a little below it at night. It is as much so on the days when the housing heats up least (on 08/10, almost all its daytime measurements): it is not the daytime sun, it is the target. On its daytime thumbnails, its spot cannot be seen (30/09 at 10:00), while those of R1 and R2 can. The compensation therefore runs every day on five references, and on three or fewer when two others are lost (passages, PA_8 and PA1_2 poorly seen in the sun, separate labels); its entries and exits do not make the targets jump (median jump of the corr no more than a little above the noise, like the other steps of the day, apart from the moments already labelled). No daytime measurement after 15/10: no return seen, until the end of the batch. Cause not established: a reflector too weak for daytime (small, dirty or poorly oriented). Ignore the curve of R3 by day; act on the target (clean it, reorient it or replace it).", "review": "2026-11-17", "analysis": "WILDSTRUBEL_2026-09-07T0000_2026-11-01T0000"}
|
||||
{"id": "WILDSTRUBEL-199", "start": "2026-10-15T08:00Z", "end": "2026-10-31T20:15Z", "family": "No data", "description": "No measurement for any sensor for more than two weeks; the metrics and the images are also missing almost all that time (two isolated values on 25/10), and come back only on 30/10 in the evening, the battery well charged but not full; it is full only on 31/10 in the morning, with daytime images on 31/10, still without a displacement measurement until that evening. Cause not established (station stopped). The curves are missing: to be checked with the operator.", "analysis": "WILDSTRUBEL_2026-09-07T0000_2026-11-01T0000"}
|
||||
{"id": "WILDSTRUBEL-200", "start": "2026-10-31T20:00Z", "end": "2026-10-31T20:30Z", "family": "Displacement", "description": "A single measurement after three weeks without measurement: the camera has nodded down a lot since 09/10 (its tilt well above that of early October, already slowly rising from 03/10 to 09/10), all the raw positions have risen in the image. The compensation runs on five references (R2 rejected), but almost all targets have their corrected curves shifted upward on corr_y, PB1_2 and PB2_2 the most, PA_9 on corr_x, and the reprojection errors of the references are higher than usual. No image at that hour. Cause not established: the camera has tilted, and the compensation does not quite catch up with it; a single point does not say whether it is lasting. To be checked: look at the mount of the camera, keep watching the next measurements.", "review": "2026-11-17", "analysis": "WILDSTRUBEL_2026-10-20T0000_2026-11-01T0000"}
|
||||
{"id": "WILDSTRUBEL-201", "start": "2026-10-31T20:15Z", "end": "2026-11-01T00:00Z", "family": "No data", "description": "No measurement for any sensor until the end of the batch; the metrics stop at the same moment, although the battery was full and the voltage normal. This is not an end-of-file delay: nothing arrives on the following days either, when the other stations have their data. Cause not established (station stopped again). The curves are missing: to be checked with the operator.", "review": "2026-11-17", "analysis": "WILDSTRUBEL_2026-10-20T0000_2026-11-01T0000"}
|
||||
{"id": "WILDSTRUBEL-176", "start": "2026-10-06T19:45Z", "end": "2026-10-07T07:00Z", "family": "No data", "description": "No measurement for any sensor during the night, except three isolated measurements at 05:00, 06:00 and 06:30, where several targets are rejected. The station shuts down for lack of energy: just before the outage, the battery is at its lowest and the voltage below its usual night level, and the metrics stop with the measurements; it restarts at sunrise with the solar charge, the compute board still cold (metrics). The curves are missing; nothing moves on the structure, but the station does not have enough energy to last the night: to be checked (battery, panel).", "review": "2026-11-17", "analysis": "WILDSTRUBEL_2026-10-06T1945_2026-10-20T0000"}
|
||||
|
||||
Reference in new issue
Block a user