supervisor chu: can you tell me from this trend -- we have four months of data here. clearly, taking a look, going from 175 to 105 to 97. is this cyclical? did we see problems one year ago? how do we understand these numbers in terms of whether or not muni is doing any better? >> clearly, the trend on page 6 is going in the right direction. the reason is going in the right direction is because service, over all, is more reliable. as the service becomes more reliable, you do not have the need to make adjustments to service.

if everything runs like clockwork, you do not need to make adjustments. to the extent there is a relationship between the number of switchbacks and service reliability, it vehicles are out, the runs are filled, there are no delays caused in the subway or in the surface -- on the surface, then the need to use switchbacks is less. if i am understanding the question, yes, the trend is going in the right direction because it indicates the service is more reliable. we have seen that as a result of the actions are described earlier, the throughput of the subway, the number of trains that make the trip, has been going up dramatically in the past three months. supervisor chu: i appreciate the comment about identifying the problem in september. on page four, you talk about

steps to improve the reliability system. what i would like, after this meeting, if you folks could work on that, is an understanding of where our lrv switchbacks have been. we cannot look at this at a four-month vacuum. we need to see if this is a trend that is going down or not. supervisor avalos? supervisor avalos: it is more a statement so that people know my point of view. i would like to think that you also agree with this. we live in the sunset, excelsior, and we are in a transit-first city. we promote that idea. but for neighborhoods that are more outside the central core,

transit first city is an empty phrase. the policy is we are going to serve the downtown core because of its density. people who live up further from the downtown core have to deal with switchbacks. it is not something easy for us to accept. i want to be able to promote transit first city but we have to make sure how we can get the sun set, excelsior, and omi have the resources necessary to make a transit first city. it has to be a transit first city in the core and elsewhere. if we cannot prevent people from asking people to get out of their cars and on the bus, we cannot expect them to do it if they do not have the service. i know we are dealing with real problems of the budget, but that is something i want to be able to make sure of. i want to work with members of

the board of supervisors to ensure neighborhoods that are further away from the downtown core can have better options, so we do not have to do with this all the time. supervisor chu: thank you. i appreciate your comments. it is clear, when we hear the fact at the mta needs to switch back and move the trains and other direction to make sure the rest of the system is working well, but for residents waiting to get home, that is not an acceptable answer. thank you for those comments. supervisor avalos: i think it is important that our neighborhoods have a say in how muni is run outside of the downtown core. >> thank you. supervisor chiu had asked for a break out of the switchbacks by time and day. in the upper right-hand side --

bare with me. i would draw your attention to two numbers. 41% and 21%. what that shows is the majority of switchbacks are being made after the morning rush-hour and then second after the evening rush hour. they are done that way to rebalance the line. the next slide shows you the condition we require before making a switch back. the point here is, what the information shows, after rush hour is when we make the majority of switchback in order to get ready, 1st between 10:00 and 1:00 in the afternoon to get ready for the afternoon rush- hour, and then after 7:00 at night to get ready for the later night service, which is important. as the length of time between trains increases, it is

important to have a reliable service at night. this was a layout of where we do the switchbacks by time of day. we also want to talk about alternatives to doing switchbacks. while a commonnly used -- while a common technique used in transit, there are other techniques used across the country. briefly, on page 8, skipping stops, bart does that all the time. they bypass patients. it is impractical for us to do it. that is a technique that works and other places. it would not be effective here. express trains are common in other places that have a different infrastructure.

probably not practical here. same reason why we do not skip stops and why dwe do not have te infrastructure. reassigning trains. when you have a delay, let's change it this train. all that really does is shift your problem. again, limited effectiveness. not a dusting service. again, the problem there is if you allow the problems to continue, you want people to go in the opposite direction. you have this bunching and you create long wait times in both directions. it inconveniences' people. one technique that a lot of people use, the first one is on top, stand by trains. in my opinion, it could be

affected here. it would be an expensive proposition but would require putting a train and operator and supervisor in a street location to plug in if there was a particular problem. we have to manage the fact that most of the places where we could put trains are in residential areas. that is something that could be plugged in. it would make a difference in improving service, albeit an expensive one. as i said last week, one of the problems we see is the level of services, with demand for our services high, we would like to have more services on the lines that go out, the n, l lines that

have heavy ridership. we have considered a number of actions. this slide on page 9, if we do nothing, right now the best option is to adjust service and to use these which backed on occasion -- switch back on occasion. failure to do so causes an imbalance on the line. on the next page, it is important. we have a recommended procedure under which we would conduct a switch back and are conducting switchbacks. there are essentially three groups that participate in managing service, the control center, line management center, the condition of all three of those are all coordinated and work together on a decision.

a first, they determine on how big a gap is, the direction, and then the best location. the other points to be made, we should not and will not do switchbacks unless there is a train in near proximity. if we are asking somebody to get off, there needs to be a vehicle right behind it. supervisor avalos: what is near proximity, and is that time or space? >> as a guideline, we say five minutes. that depends on a couple of factors, like the location of the delay, severity of the delay. depending on what causes the delay, we could have more than one train delayed. >>supervisor avalos: so what is

the space generally in between -- >> there are two things that have to happen. five minutes is usual. actually, three things. there needs to be a train five minutes behind. no one should wait more than five minutes. more importantly -- supervisor avalos: that is a good rule, but my question is, what data do you have, what is the proximity you are using? what do you generally range from? the experience people have is it is beyond five minutes. if so, we should be able to measure that to improve service. >> i do not have that with me, but i can certainly get it. i am happy to bring that

information to the committee. we will forward that as well. supervisor chu: we have seen from many of the folks, it even if what the mta is verbally telling us that they will not make a switch back unless there is a serious delay, that there is also a train five minutes behind -- the experience of riders is the train is not five minutes behind. we had an incident in december were two trains were turned back. for that individual, they were certainly not winning just five minutes. the question that supervisor avalos asked is relevant. although the policy is five minutes, do you have a policy that no consecutive trains are turned, a measure that that is the policy being followed? that is the experience being seen by residents.

supervisor avalos: she asked you a question that you should be responding to. supervisor chu: the question is , we have a problem with the mta putting out statements that we are turning around trains with significant delays and that there is one with five minutes behind. there is no way to know that is true. from everything we see, we see trains are being turned around consecutively. we hear they could be coming 20 minutes later or even later. what is your experience of what writers are experiencing is what you say? you have data that shows us that the next train should be coming by. >> so the question is, do i have statistics on when switchbacks,

to show you when there was a train -- how long people waited and how long the train was? i do not have it with me but we can get it. supervisor chu: supervisor avalos? supervisor avalos: i represent people outside the city. i feel pretty angry about how unprepared you are and how you are going through these questions and how you are not really ready to respond to these questions. it is not like you are listening to us. i do not generally get agitated like this. i know there is a sense of agitation from riders. i wonder if we should even continue this spirit we are not getting the answers we want. i would like to have a change in how you are giving your presentation so we can have a better sense. if we ask a question, you should be able to give an answer back.

what i am hearing is unacceptable. supervisor chu: thank you, supervisor avalos. one of the things we could do, there are some slides that are at i am interested in getting to which includes the number of switchbacks on individual lines. perhaps we can go to those and have more directed questions for the mta. would that be ok? ok. >> on page 11, those are the characteristics that you know of the lines. on page 12 is the n line. the point of this is to show, over the course of a year, the number of trips we run nurses

the number of short term that we make. it is a small percentage of the total trips. the other point to make, when we talked about wanting to inconvenience the fewest number of people, switchbacks are in generally done at sunset. trains are turned to support, through the greater good of riders, and minimizing inconvenience to people. we recognize when someone is asked to get off, they are inconvenienced, even if there is one behind it. we do everything we can to minimize the impact on the least number of riders. similarly, on the l line, a small percentage of turns.

the same principle near the end of the line. you inconvenience to the least number and your balance the line for the greater number of people. m line, same percentage. we recognize anyone that is asked to get off a train, there is an inconvenience. the passenger complaints. over the course of two years, as you have noted, we have had over 150 complaints at 311 on switchbacks. a majority of the complaints have to do with our inconsistent communications. people need to know as early as possible if there train is not going to the end of the line. then there has to be consistent announcements, both by the operator and within stations, to make sure people are not

confused. we have taken a look at the data. this is a sampling by line. one of the things that we have done since we first spoke to you about this in september is we have made a change in the software so that the destination sign, the voice in the station, the station platform, all have the same destination. that had been an inconsistency in the system. that gives us the ability to let someone know early on if there train is going to be switched back. one of the things i am laying out on action items is to continue to make sure that our sign-ins' it is proper, consistent on announcements, and that we follow the policy that there will only be a turn back if there is a train five minutes behind. finally, in terms of the

overview from our perspective, our goal is to provide the most reliable service without the need to switch back. we are taking a number of actions to improve the reliability of the system. we are certainly willing, and are investigating other steps, particularly on the surface to find out how to get more reliable. supervisor chu had asked for the cause of delays in the meeting last week. earlier this morning, we provided some of that which shows about 50% of the delays from november to december were caused by muni, about 50% caused by other factors. from our side, we have to make sure we work on our vehicles affectively. that was a pivotal source of the delay. we look forward to working on this. the other kind of thing we can

work on, talk about, with your help and support, before, parking lanes, traffic enforcement. we want to run the most reliable service which should result in the fewest number of switchbacks. that is our overview. supervisor chu: can you explain to me why, looking at these different -- i know supervisor avalos talked a little bit and we talked about the problems we have with our buses. today we are just talking about light rail vehicles. i know there are further problems with the reliability of buses. on these rail lines, we see different percentages of how turning is affecting people. on the m, we have roughly 1% turned. in terravale, it went from 1% to

3%. in judah, it went from 3% to 6% of trips being turned. why is it that the end judah has seen this number of switchbacks? you know that that is particularly problematic. there are so many writers on this line. -- riders on this line. >> there were two questions. one about the reliability of the bus fleet? supervisor chu: we are concentrating on the light rail vehicles. i fully appreciate problems to the other bus fleet. with these three lines, we have different numbers of prevalence where these turns occur. you have them on slide 12 through 14. we see anywhere from 1% going up through 6% of trips that are

short term. the biggest impact is on judah where we have 6% of trips being turned. why is it that this occurs so unevenly across these different lines? >> so the percentage of trips that are turned on all three lines is under 1%. in terms of the n judah, it is one of the most heavily traveled lines. that is basically the reason it can be turned more frequently, if that is the question. again, i am not sure i understand the question. supervisor chu: we see much more

prevalent on this line than others. why is that? the n judah had 370 turns in 2010 compared to the l terravale, 177. >> just for introduction purposes, i am jim kelly of rail services. it may be a misunderstanding on our part. if you look at the n judah, that is 0.32%, not 3% of the cars being turned. to speak regarding the n line, the n judah has the most service of any of our lines and also has the most flexibility. the k is probably the best advantage. once the train leaves st. francis circle, there is no

opportunity from there until balboa park to turn. there are no switchback locations on ocean avenue. so a lot of our adjustments on the kt line have to take place on the t line because that is where we have the greatest ability. if we do not take any action on the t line, there is no way to adjust the service physically coming back the subway from embarcadero to the remaining trips on ocean avenue. on the m line, the switches at 19th avenue were in a state of disrepair. they went back into service after the st. francis work. the only location where we could switch back was at broad and the imf, roughly eight minutes from the terminal at geneva, which affected the seven or eight

stops. in a lot of cases, what we had to do was make adjustments at the other end of the line at embarcadero by extending trains out to fourth and came, or do what john mentioned earlier, reordering trains. it was necessary to balance the lines. so to your larger question as to the n line compared to others, it has more service. so the opportunity to service is greater. there are 17 trains operating in rush-hour on the n judah, compared to the l or m. so the percentage is larger just by sheer numbers. supervisor chu: just so the public is aware, the slide shows 0.03, but that is percentage. supervisor avalos? supervisor avalos: i am just

someone unfamiliar about the n judah line. you have to walk a great distance through the sunset to get there. you are waiting longer and you already have to walk a long way to get to the n judah. the options in the far parts of san francisco are less rich than elsewhere, so the inconvenience is greater. i will keep emphasizing that for years. >> to speak to that, there is one pilot program that we are looking at in the outer end of the sunset. we have had conversations with supervisor chu. that is the possibility of an

express bus. that would operate from all playa, operate at local stops up until 19th ave. then it would operate as an expressed vehicle, work its way down to the financial district on an approximate 10-minute headway. we are looking at potential for buses that would have wi-fi to make it a little more convenient for people. it has a twofold advantage for us. it's the alderson said people a direct link to downtown very quickly -- it gives the our sunset people a direct link to a downtown very quickly. also, those passengers who decided to take the express bus, it does great capacity for existing fleet because as much as we would like to put additional vehicles on the line, we do not have them available at

this time. this would be a work around, and this would operate in the a.m. and p.m. peaks, so five days a week. supervisor chu: with regards to one last issue that i do not think was really address -- at the end of the day, in order for the system to be more reliable and for switchbacks to not be necessary, we have a number of reasons why the delays occurred to begin with, and we have not had a very in-depth conversation about why it is we see these delays occur. we talked about system failures, potential problems with the equipment, trains being blocked because there are cars on the rails -- what kind of information do you have to explain what the root causes of the failures are? why are we seeing these delays in a system that you believe necessitate a turnaround? how can we change those root causes of the problem?

>> i think john clearly explained the mechanical problems we are experiencing. we analyzed all the data to what the causes are, the root causes of vehicle breakdowns. they tend to be propulsion problems. we had a unique system with steps that go up and down depending on the high platform, the low platform. our door mechanism systems have over 220 moving partners -- moving parts, so it is a very complex door, and like a barn door, which is essentially a hinge and a door. when you have that many moving parts, there is chances of failure we're looking to each of these components to see how we can improve the reliability. a five-minute delay on the subway seems relatively short, when you are operating of words of 50 trains per hour through a given