[LEAPSECS] leapseconds on trains (was Re: No leapseconds on trains)

Rob Seaman seaman at noao.edu
Thu Nov 17 09:18:58 EST 2011

On Nov 17, 2011, at 4:20 AM, Poul-Henning Kamp wrote:

> We're having a bit of a project management scandal in Denmark related

> to purchase of 83 "IC4" trains.

I suspect I'm not the only American reading this wishing more of our scandals were about trains...

> Reasearching this, I have been reading up on MVB, "Multi Vehicle

> Bus" (IEC61375) which is how modern rail-hardware talks to each

> other, which is good geek material btw, some smart thinking in

> there. [...]


> I wonder what would cost more to fix, test and recertify ?

Absolutely nothing about the ITU process to date has focused on answering such questions. A suggestion of using systems engineering best practices to grapple with the issues is met with snores or a snort of derision. How do you think project management scandals get started?

> A) The majority of rolling stock built in the last 10 years

> or

> B) A few astronomical telescopes.


> Actually, I don't wonder, I know the answer to that one: You can

> build several ELT's for what A) will cost.

No, you don't know the answer to this. First, your argument assumes that A and B are an exhaustive inventory of the issues. Second, you are artificially linking issues caused by a failure to understand and follow a standard that has been in place for four decades with issues that would be caused by redefining that standard. The astronomical telescopes represent projects and engineers who did follow the standard.

Third, a Y2K-like inventory of UTC dependencies would also factor in the *severity* of the current leap second issues (apparently small) versus the future UTC-is-no-longer-UT issues (clearly very big). Fourth, such an argument assumes that the trains themselves don't have dependencies on UTC remaining UT. (How many clocks are on trains? How many systems other than MVB must interoperate? Rail transport interoperates with other types of transport, shipping, trucking, air. Etc.)

Fifth, even assuming B is limited to astronomy (it isn't), "few" means hundreds or thousands of professional astronomical telescopes and many more amateur community facilities - and "telescopes" aren't the only astronomical systems affected. But those telescopes include multi-billion dollar assets such as the Hubble, and unique and irreplaceable facilities such as neutrino telescopes beneath the Antarctic ice ("embedded technology", both literally and figuratively), gravitional wave telescopes with their own smart-thinking networking protocols, and solar telescopes whose data are used by the Air Force to generate space weather reports pertinent to the safety of the spacecraft fleet. You might value astronomy less than Danish trains. Even from a purely economic standpoint an unbiased accounting may disagree.

However, the Exton meeting (http://futureofutc.org) was jointly sponsored with the American Astronautical Society and the American Institute of Aeronautics and Astronautics. (Thanks for the suggestion to invite the corresponding railway associations to contribute to the next meeting.) How does the economic value of rail compare with air travel and space operations? (And navigation at sea and trucking?) What are the risks, joint and separate, of redefining UTC? What are the *actual costs* of updating duplicate instances of standardized railway hardware and software - in locations that programmers are guaranteed to be able to reach by train - versus updating numerous state-of-the art systems pursuing unprecedented research at unique facilities located literally at the ends-of-the-world (and above it and beneath it).

(Hint: the non-scandal-causing process to find out is *not* to "wing it".)

...and sixth, 10-year-old rolling stock would have encountered at least two leap seconds. There should be plentiful real-world data on the risks and costs. But by all means let's make the project management scandal bigger.

> For instance, as part of the validation of the *concept*, a special

> train was run in passenger service for two years, at a total cost

> of over 3M$

And you don't think they should invest even one penny to understand potential dependencies they have that Coordinated Universal Time remain a type of Universal Time?

I don't know the answer to whether it would cost more to fix any speculative problems resulting from failure to follow the current standard versus the definitive problems that will result from redefining that standard out from under those who did follow it. But I know enough to ask the question and that the figure of merit includes the contingent risk assessment. Good project management would involve actually performing the obvious inventory, not speculating.

> But the really interesting thing to remember here, is that if you

> "asked the railroads about leap seconds", what are the chances you

> would get somebody on the other end of the line, who knew that the

> MVB standards would have to be revised, and _all_ compliant devices

> have to be reworked, retested and recertified to the new standard,

> in order to *continue* leapseconds ?

(Fallacious reasoning: http://philosophy.lander.edu/logic/complex.html)

Even *if* it proved hard to find a railway engineer knowledgeable about timekeeping, the rest of your assertion is assuming contingent answers not only to that question but to a host of others. But we're all familiar here with the history of standard timekeeping and know that the railways played a special role. You may assume that timekeeping is beneath the rail community's radar, others wouldn't.

Speaking of radar - there's another good example of 1) a technology dependent on Earth rotation, and 2) that has pertinent real world experience handling leap seconds.

And of course the "new standards" MVB will have to accommodate would include the redefined UTC. Just because MVB or any other system might not explicitly handle leap seconds doesn't mean that operational practices of the systems layered on MVB don't do so. In fact, trains in Denmark and elsewhere are guaranteed to accommodate leap seconds implicitly. All the clocks change one way or another. Stop introducing those adjustments to keep UTC ticking mean solar time and their absence will move the entire Danish rail system to the equivalent of TAI. You can't imagine this would cause any problems. I can. Actually looking is the way to find out.

> As much as we may think of the leap-second debate as a technical

> issue, it is primarily an economic issue.

Set up the straw man and knock it down. Has anyone here not been viewing this as an economic issue? But does proper accounting really convey some sort of argument (in dollars and cents) that the appropriate inventory of UTC dependencies should not be performed? Would a neutral observer agree with a suggestion that the better project management choice is the equivalent of a tight rope walker wearing a blindfold?

Leap seconds are a sideshow. The big top will be the problems caused by redefining UTC to no longer provide actual Universal Time.

Rob Seaman
National Optical Astronomy Observatory

More information about the LEAPSECS mailing list