GPS Week Number Rollover Apr 6/7
Expect some amount of time related disruption around April 7 00:00 GPS as the GPS 1K week cycle rolls over and GPS receivers that have not been patched or updated to use new CNAV and MNAV messages, or compensate for not doing so, provide incorrect time, position, and navigation solutions. Garmin, for example, rolled out updated firmware for some of their receivers, including the 18x popular on NTP servers, just over a week ago. Previous rollovers have shown symptoms of clocks being out by multiples of 256 weeks, some resetting to the GPS epoch at 1980 Jan 06 Sun 00:00:00+0000, showing misleading position and navigation solutions, depending on what inputs are used to solve the 4D simultaneous equations. Airlines, ships, and well maintained commercial trucks shouldn't crash, but the rest of us should check any devices using GPS! If my script is correct, the rollover occurs at the indicated instant on the time scales below: 2019 Apr 06 Sat 23:59:42+0000 JD 2458580.499792 MJD 58579.999792 ToD 86382 NTP 3763583982 Unix 1554595182 GPS We 2048 Cy 2 Wn 0 Wa 0 ToW 0 DoW 0 ToD 0 TAI-UTC 37 TAI-GPS 19 GPS-UTC 18 [*NOT* an April Fools post, unfortunately!] -- Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada This email may be disturbing to some readers as it contains too much technical detail. Reader discretion is advised.
On 4/1/19 1:28 PM, Brian Inglis wrote:
If my script is correct, the rollover occurs at the indicated instant on the time scales below:
2019 Apr 06 Sat 23:59:42+0000
That hits a bit close to home, as I'm scheduled to take a US commercial flight that will take off about two hours after that. As it happens in my software engineering class today I talked about AeroData's nationwide outage this morning, and now you're prompting me to speculate whether that outage was related to the GPS rollover. Although AeroData hasn't said publicly what caused the problem, the Orlando Sentinel reports that it was related to GPS and it's not beyond the realm of possibility that a flight-planning application would use prospective GPS clocks that wrapped around. (Of course I'm just guessing and could easily be quite wrong.) See: Pedersen JM. Nationwide glitch impacts flights at Orlando airport; Southwest, JetBlue most affected, official says. Orlando Sentinel. 2019-04-01 08:05 -04. https://www.orlandosentinel.com/news/breaking-news/os-ne-oia-aero-data-glitc... The outage caused hundred of US flights to be delayed today. See: Gallagher S. Flight management system crash causes airline delays across US. Ars Technica. 2019-04-01 16:34 -00. https://arstechnica.com/information-technology/2019/04/flight-data-system-ou...
On 2019-04-01 17:41, Paul Eggert wrote:
On 4/1/19 1:28 PM, Brian Inglis wrote:
If my script is correct, the rollover occurs at the indicated instant on the time scales below: 2019 Apr 06 Sat 23:59:42+0000 That hits a bit close to home, as I'm scheduled to take a US commercial flight that will take off about two hours after that.
As it happens in my software engineering class today I talked about AeroData's nationwide outage this morning, and now you're prompting me to speculate whether that outage was related to the GPS rollover. Although AeroData hasn't said publicly what caused the problem, the Orlando Sentinel reports that it was related to GPS and it's not beyond the realm of possibility that a flight-planning application would use prospective GPS clocks that wrapped around. (Of course I'm just guessing and could easily be quite wrong.) See: Pedersen JM. Nationwide glitch impacts flights at Orlando airport; Southwest, JetBlue most affected, official says. Orlando Sentinel. 2019-04-01 08:05 -04. https://www.orlandosentinel.com/news/breaking-news/os-ne-oia-aero-data-glitc... The outage caused hundred of US flights to be delayed today. See: Gallagher S. Flight management system crash causes airline delays across US. Ars Technica. 2019-04-01 16:34 -00. https://arstechnica.com/information-technology/2019/04/flight-data-system-ou...
From the Ars commentary, I'd be worried about any airline using date, software, or services from AeroData Inc. for flight performance, weight, and balance planning, given the apparent lack of attention or care to maintenance of their sites. If there is any "GPS" component, I would expect that it because of the North American tendency to use GPS to refer to satellite augmented navigation systems (SatNav in Europe) or any systems using GNSS in general. In this case it might refer to provision of navigation information, calculating and generating waypoints and markers for flight tracks, with runway, heading, gradient, temperature, beacon, marker, control frequency, fuel load and usage, weight, and balance annotations. Their name appears in a case study by VMware: they apparently didn't use it effectively regardless of the cause of their issues.
One frequent flyer posted "Delta mainline (and American) was not hit because they don't use their takeoff data from AeroData. Delta Connection as well as Southwest, Frontier, United, and JetBlue were impacted". They don't appear to be the kind of people from whom you could expect a good pull request, or one which you might care to accept. ;^> -- Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada This email may be disturbing to some readers as it contains too much technical detail. Reader discretion is advised.
participants (2)
-
Brian Inglis -
Paul Eggert