« Who Should Run San Francisco and Muni? Ask The Muni Rider Voter Guide | Main | The Next Time You're Doing the BART/Millbrae Station Dance Remember This »

Friday Roundup: An Event on Irving, Outside Lands, Muni Rider Voter Guide, Etc.

Laundryshowpostcard-1.jpgThere's about half a dozen pieces of blogging detritus hopping around on my Mac's hard drive right now and not enough time to write a separate post for each one, so here's a short roundup on a number of topics.

First, however, is an art show I'd heard about last night at the Inner Sunset meetup at the Little Shamrock. Entitled "Dirty Laundry," it's an art show hosted by ART I.S. at Laundrapalooza (an actual laundromat), on Irving between 23rd and 24th. It's an interesting idea, to say the least and if you're in the neighborhood between 7 and 10 and aren't going to Outside Lands, you should check it out.

- Speaking of Outside Lands, this year's music festival starts tomorrow, August 14th and will last through the weekend. If you're coming out to the festival, don't be a fool and drive. THERE IS NOT GOING TO BE ANY PARKING IN THE NEIGHBORHOOD AND TRAFFIC WILL SUCK. Both the Richmond SF blog and Akit's Complaint Department offer some suggestions on how to enjoy the festival, or at least not have it ruin your life if you live nearby. Personally, I think by only having it on the weekend, and leaving out Friday made it easier for everyone - the Friday commute isn't hosed and it sounds like this time around things will work out better than the first one a few years ago. Plus they have locals providing the food this time. Cool.

- My side project, the Muni Rider Voter Guide, received its first candidate questionnaire returned today. Candidate Jim Meko, running in District 6, sent in his responses just a day after receiving the questionnaire in the e-mail. I was curious to see who would do so, and decided to have a secret prize for the first respondent - a pre December 2009 cuts map of the Muni system which I'd obtained at the end of 2009 from the MTA via then-spokesman Judson True.

Not only is it laminated, in mint condition and easily framed, it is also a reminder that this recent chatter about "restoration of cuts" is not even close to the cuts they made back then - and labeled an "overhaul." Congratulations Mr. Meko and candidates, please send in your questionnaires so voters know how you stand on Muni issues.

-Another update: You may or may not recall a challenge I made to Muni management to actually ride Muni. In June, I marked off Day 42 of the challenge, having heard not a word from said bigwigs. I know, shocking, right?

Well now you can tack on another 45 days to that, bring us to Day 87 of no response from Muni management. You can also tack on 45 days for the long awaited answers regarding the Muni shelters on 9th and Irving as well. Tick tock, MTA!

Finally, this really doesn't have a lot to do with anything, but I collected the "The Social Network" movie trailer (My GOD that movie sounds like it's going to be snoreville) along with all the parodies, which are way more entertaining. Go to the Blog Some People Read, and check it out.

Comments

This doesn't really have anything to do with anything, but I was actually on the N Judah today, and I must say it was rather disappointing. There was a very large gap in service going outbound, and they didn't even bother to try to fix it going inbound (for example, by sending a train express). There was slowness, bunching, and general inconvenience. And best of all, that first train then spontaneously decided to turn back at Embarcadero, dumping out me (I had just missed my Caltrain) and a number of displeased Giants fans. Is there actually some kind of dispatcher somewhere to manage the service, and if so, why do they do such a poor job? Also, why does the stupid signal system not allow double berthing? The Market Street platforms are so nice and long, and dwell time is the limiting factor in any local-stop transit operation, so it would make far too much sense to allow that sort of thing.

1) Central control for the Muni's Metro is a semi-bunker next to the West Portal Stn.

2) There is a dispatch booth at Embarcadero Stn. and possibly another one at Duboce + Church. A field dispatcher is sometimes stationed at the control box at Fourth + King. A roving inspector can be seen at various terminals (look for a guy in blue uniform w/ radio + clipboard) from time to time.

3) The block signal system for the Metro is a coarse-grained, probably cheese-pared, near-kludge. While sensor and computer technology evolve at a mind boggling pace railroad systems proceed at a pace worse than mil-spec. This is due to budgets, federal regulations, and operational barriers. Essentially, the system would have had to be built in duplicate form to allow a smooth migration from one control generation to another. That kind of double-piping would, in most shops, be rejected from the get-go.

4) Platform length and consist length are long standing bones of contention. The short answers are [a] the Market St. stations were built to BART standards and [b] the Metro is a hybrid system with TWO (2 !!) speed regimes. Various factors (dwell time surface vs. subway, surface speeds, subway speeds, platform layout, etc.) are combining to put the Metro in an expensive trap. The only way out may be a total rebuild of the system to allow low-floor cars that are double-articulated.

Unfortunately, such a rebuild would cost an obscene amount of money and time. Look at the price tag on the Chinatown Subway project. That poor turkey is going to face marketing problems due to missing / mis-located stations and will probably trigger a rider revolt when they cut bus service on the surface.

Ted K: the signal system was already upgraded to the fancy computer-based one you're talking about. That seems to be the source of many of the problems. They bought the awesomest latest technology based on the awesomest operating system available at the time: OS/2. Now they're stuck with it. At least when you have relays, you can re-wire them yourself. When you have a binary blob, not so much. As for platform length, there's no need to have low-floor double articulated cars. You can have three or four car trains (and they did at some point, maybe before the current signal system). You can also use the full length of the platform to stop two trains at a station at the same time, as is done in places like Philadelphia and Boston, which increases throughput somewhat, since dwell time is the limiting factor in subway throughput. But again, they can't do that due to the signal system, which they upgraded from the "coarse-grained near-kludge" to the shiny new computer age of OS/2.

Anonymouse - The present system IS the coarse-grained near-kludge. Remember the nasty West Portal Stn. accident ? That was caused in part by an operator manually overriding the system. The safety margins are very conservative. Those same safety margins are probably also one of the factors in not allowing controlled multi-train stopping. Do you want more manual operations ?

Also, assembling a multi-car consist from two different lines can add delays to the system. Plus, I suspect that the track configuration under Market St. may have a subtle capacity ceiling due to [a] double-tracking (vs. triple- or quad-), [b] the Duboce + Church fork, and [c] the Embarcadero Mini-Yard. Those three elements constrain the system when everything is working properly. But if the control system hiccups in those areas [b + c] that fragile balance gets strangled. I would like to see a report on [1] the amount of slack in the timing of the runs from Van Ness to Church and [2] what's the trains-per-hour capacity of the mini-yard with none to a pair (0, 1, 2) of glitches.

The reason I see a low-floor, double-articulated car as a dream solution is due to boarding problems and platform usage. Climbing up and down those steep stairs on the Breda cars makes the PCC cars seem almost friendly. Part of it may be due to the railing configuration. A double-articulated car would add capacity and use two-thirds or so of the Market St. platforms without narrowing the safety margins.

P.S. Control / signal system design points :
Wire plan - home-run vs. nodal vs. overlapping nodal;
Sensor mounts - single-, double-, or triple-head;
Control system interface - hard-wire vs. custom buss vs. ARINC-429 / AFDX / Ethernet.
An ideal control system would have [1] a master sensor node at each station that reads sensors all the way to the next station's platform, [2] triple-head sensor mounts with two heads in use and the third available for upgrading, and [3] use a common, proven buss technology that is widely available. (ARINC-429 busses are used in airliners.)
http://en.wikipedia.org/wiki/Avionics_Full-Duplex_Switched_Ethernet

P.P.S. I'm a long-time computer engineer with experience in databases and networks. Part of that experience is a decade in the aviation industry in a support role.

P.P.P.S. Ugly question - How many fiber optic conduits are there in the Market Street Subway (MSS) / Twin Peaks Tunnel ? If the answer is none then people from SFMuni to the Office of the Mayor have dropped one hell of a ball. If the MSS was designed without conduits and tap points then the ball dropping goes all the way back to the 1960s. Had those conduits and tap points been designed in and built then SFMuni and BART could have been getting rent from AT+T (ex-Pac Bell), Comcast, etc.

Ted K: the only reason I even suggest double-berthing is that the current automatic control system will, when there is congestion, drive a train right behind the one in front of it, such that the back train is fully within the confines of the platform. The system won't allow it to open its doors, though, because the system considers a station to be a point, rather than a segment.

Anyway, if you want the gory details of how the system works, it's a Seltrac system. There are a inductive loops running the length of the track, with the wires crossing over every 20 meters. Each loop is much longer than that (I think about as long as the distance between stations). These loops are used both for locating trains and bidirectional communication. The train detects its position by counting these crossovers and reports that position to the Central Control Computer, and the Central Control Computer tells each train how much further it can travel, with the on-train computer setting a target speed based on that.

The older, simpler, and probably more reliable system uses coded track circuits for continuous train detection, wayside-to-train communication of target speed, and communication of track occupancy to block sections in rear. Railway signalling has a long history, and while I'm sure there's much to be learned from aviation, it's also a rather different environment in a number of ways (for one thing, you're communicating to a moving train in a tunnel, not inside an airplane). If you want to discuss the design of railway signalling systems, though, I'd be happy to do so.

Greg... SHARP is back in business, at 9th Ave, between Moraga and Noriega, 1738 to be precise. Our new H.Q. bld opens with a Dedication ceremony tomorrow...
Politicians and snacks and a tour of our new 4 story bld....


7pm to 8.30... We want to rebuild our member base, to serve as a community center, more than ever.
See www.sharpsf.com

Cheers,

jb

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)

GOT A STORY TO TELL?

Drop us a line and share your tales of MUNI woe, City life, and more with your fellow citizens and MUNI riders!






Stay Updated




Powered by FeedBlitz

SITE CREDITS

© 2005-2011 - Greg Dewar - All Rights Reserved

Art by Quellette Design

Powered by Movable Type 4