AC Transit’s ideas for making the 51 faster and more reliable

17 Mar

Here’s part two of my posts about the 51 open houses and planning process. If you haven’t read it yet, I recommend first reading the previous post on the reasons the 51 is so slow. These recommendations probably won’t make much sense otherwise. Also, if you’d like to learn more, read the entire AC Transit 51 report.


Stop removals and re-spacing: As explained in the previous post, the stops on the 51 are too close together in most places. To fix this, AC Transit proposes to remove 22 bus stops (15% of the total stops) along the line. The estimated time savings would be more than four minutes in each direction. This estimate sounds conservative to me, but, regardless, I think this plan makes a lot of sense. The only problem? “AC Transit does not own or maintain bus stops, and so the local jurisdiction controls the final placement of the stops and works with AC Transit to determine the best location.” So ACT needs to go through the planning processes in Berkeley, Oakland, and Alameda to get these stops removed. And at the Berkeley open house, Berkeley Councilmember Jesse Arreguin raised concerns about removing the stop at Milvia on University. I’m concerned that community members will rally around saving several specific stops, and without removing a significant number of stops, there will be no significant time savings.

Make boarding and alighting faster: ACT recommends several ways to do this. One is lengthening bus stops – right now, many bus stops are too short so buses have difficulty pulling in and out and passengers sometimes have to board and alight in the street, which is not only dangerous, but also slow. A related idea is to clear the areas around bus stops. Street furniture, trash cans, and newspaper racks can get in the way of boarding and alighting so anything that’s close to a door should be removed. To make boarding faster, ACT recommends encouraging Translink use, which is much faster than fare passes or cash. Translink is also SO much more conenient, but I’ll get into that in another post. And lastly, ACT wants to encourage rear door alighting. Yes! I can’t tell you how much time I see wasted as riders alight through the front door. Unless you have difficulties moving, there’s not good excuse to alight from the front door when others are boarding there.

Supplemental Service: Currently, ACT provides supplemental service at Oakland Technical High School and Claremont Middle School. The report recommends using articulated (double) buses at Oakland Tech and to create supplemental service for Berkeley and Alameda High Schools. Similarly, ACT wants to work with UC Berkeley to determine if its bus service is sufficient, particularly on the perimeter line that travels around campus. They also recommend creating better wayfinding signs around campus so students know they can use the perimeter buses to travel short distances. I’ve often seen UC Berkeley students often hop on AC Transit buses just for a few stops, when they easily could walk or use the permiter bus so hopefully this can be addressed.

And More:

  • Schedule Refinements: Optimize schedules so that there isn’t too much nor too little run time (this sounds a bit optimistic).
  • Signal Coordination/Signal Actuation: Coordinate signals so the buses can make it through several green lights in a row.
  • Real-Time Passenger Information Systems/Passenger Amenities: There’s no money to put NextBus electronic signs at 51 bus stops so instead they recommend putting up signs explaining how you can access NextBus online or via PDA. Too bad for those who don’t have internet access.
  • Scheduled Dwell Point: Staff recommends using Rockridge BART as a dwell point, where a few minutes padding would be added to the schedule. This might cause some buses to sit at Rockridge for several minutes but it will help others make up lag time.
  • Active Line Management: Supervisors on the route have been shown to improve on-time performance so ACT would like to have more active line management on the 51 line. Of course, this costs money.


At various segments along the route, AC Transit has proposed specific improvements that would save time. I won’t cover each of the specific segment recommendations here, but these are the types of recommendations that were made:

Queue Jump Lanes: These are lanes that are short distances approaching an intersection, often taken from parking lanes or right turn lanes, where a bus can pass a line of cars as it heads to a bus stop across the street.

Peak Hour Parking Restriction: Parking would not be allowed during peak hours so that there would be additional travel lanes in heavily traveled areas (like University near the freeway) that could be used both by buses and other vehicles.

Restrict Turns into Driveways: I’m sure many of you can imagine where this is a problem – on College near Safeway. Restricting turns in this area would not only help speed up buses but would also speed up all traffic.


And now to the part you’ve been waiting for. After looking at all of the 51 data, ACT staff came up with three proposals for changing the route service design. They kind of took the easy way out by not telling us which one they recommend, and I find it almost impossible to believe that they don’t have a favorite (though maybe they couldn’t agree). At this point, no decisions have been made in implementing these changes – they’re going to have lots of public hearings at the city and regional levels. Here are their recommendations:

Split Route: This would consist of splitting the 51 into two separate lines. These two lines would continue to serve all of the stops on the 51 route and would together cover the entire distance of the 51 route, but each would only transverse about half of the route. A split route has a couple of advantages. First, shorter lines could isolate the delays so that a traffic jam in Alameda didn’t effect riders in Berkeley, for example. Second, dividing the route provides flexibility in that the two lines could have different levels of service (ie 10 minute and 8 minute headways) or use different length buses.

However there are some huge problems with this proposal. An important one, especially in the face of ACT’s budget crisis, is that the two lines would be more expensive to operate than one line. And then of course a split route would be detrimental to those who would need to use both lines. That would involve a transfer (which may cost money) and extra waiting time, though ACT staff say this would be mitigated by having well timed schedules. To that I say, yeah right – I’d be more likely to just avoid the 51 or walk to my destination rather than deal with transferring.

The report discusses three possible points for breaking the line but staff only focused on two at the open house: breaking the line at Downtown Oakland or at Rockridge BART. A DTO split would effect 1639 riders daily who take the 51 through that point and a Rockrige split would effect 2439 riders. The report kind of brushes off this effect, since it’s only about 10% of the ridership, but that attitude bothers me. I can’t help but think that a lot of these riders will figure out some other way to get to their destination, especially if they already transfer to or from the 51 from another line. I’m not so hot on this option, and maybe it’s because either of the proposed splits would effect me on some of my 51 rides, but most of the attendees at the Berkeley open house preferred this option.

Limited Service/Local Service: This option consists of maintain current 51 service that stops at every bus stop but adding in “limited” service that would only serve the major stops along the route. I’m going to admit that this is my favorite design plan, but I wish they would stop calling it limited and start calling it rapid! I remember the 40 before the 1/1R existed and that it had many of the same problems that the 51 has. Buses would get bunched, rides would take forever because we stopped at every stop, and buses would get so crowded that sometimes a bus would pass up riders because it was too full. The 1R hasn’t eliminated all these problems (especially reliability and bunching, which is why we need BRT), but at least now buses aren’t uncomfortably full and once you board a 1R, you’ll probably get to your destination quickly.

Now the 51 line is of course not the same as the 1 line. For starters, College Avenue will never have rapid service because it’s difficult for buses to pass each other and there’s often near stopped traffic for several blocks. But for the rest of the line, particularly along Broadway and University, which have several traffic lanes, rapid buses could move very quickly. Really, this is the type of service I dream of. Unfortunately, this option got just three paragraphs of attention in the report, which makes me concerned that ACT isn’t taking this option seriously. I hope that changes.

A/B Stops: The concept of this is that there would be two 51 – 51A and 51B – that both ran the entire length of the 51 route. Most stops would be labeled either A or B, except major stops that would be labeled both A and B. The 51A would only stop at A stops and the 51B would stop at B stops. Confused yet? You’re not the only one. When staff presented this at the open house, the whole room looked perplexed. This plan seems crazy and I wonder if staff came up with it just for the purpose of having a plan that could be easily taken out of the mix and that could be used later in arguments – “At least we didn’t implement the A/B option.” Seriously though, nobody at the open house liked this idea and I’m guessing it’s not going to go anywhere so I’m not going to worry about it.

No Change: There’s of course the option to not change the service design. ACT could leave the design as it is and just make the other improvements listed above.

Whew. That’s it, I promise. Now that I’ve taken hours of my time to pore over this report and dissect it, I’d love to hear what you think. Will any of these suggestions work or is the 51 doomed to slow, unreliable service? I’d especially be interested in hearing what you think about the route level service design options or if there are improvement suggestions you have that were not covered in the report.

5 Responses to “AC Transit’s ideas for making the 51 faster and more reliable”

  1. R. Kaplan March 17, 2009 at 8:08 pm #

    Thanks for this great write-up!

    For people who want to check out the real-time bus arrival information, it is now available online (not yet for all bus routes, but a start-up program with a pretty long list of routes covered). Check it out at:


    Then select “Northern California” then “AC Transit”

    Best wishes,

  2. Just Anon March 18, 2009 at 2:42 pm #

    Awesome. You summed it up great. Curious to see what others think as well.

  3. JHorner March 19, 2009 at 8:28 am #

    Another fantastic post.

  4. david vartanoff March 19, 2009 at 9:11 am #

    was at the Monday mtg in Berkeley. here’s what I sent to AC.
    Route 51 a response

    In 1993 the Alternative Modes Analysis (DKS consultants) identified the 51 as one of the top five ridership routes and studied many options for improving service. While the 51 was not deemed a candidate for trolley bus or Light Rail, the study also described as Transportation System Management several potential steps to improve the route but assumed nothing could be done. A few stops were eliminated, overnight service abolished, and 1965 riders a day have been lost. (compare figures DKS 1991 20,712 stat from Table 6-2 Page 6-10 with 18,747 Planning Focus 2007 page 7). Now, 15 years later the new Planning Focus / Service and Reliability Study revives the TSM concepts with some further enhancements and explores splitting the route in order to localise the effects of line delays.

    Once again stop eliminations are proposed, although the delay analysis makes clear that 80 % of delay time is traffic signal generated. Signal preempts were rejected in the DKS study as merely shifting traffic jams to other streets—a policy sometimes called transit first. AC has only deployed signal preempt hardware on Rapid buses even though local buses provide the majority of service hours on Rapid routes. Unfortunately some of the most persistent delay hotspots are currently involved in redesign or redevelopment projects beyond AC’s control (College Alcatraz to Claremont, the area of the downtown Berkeley BART Station for instance).

    Dwell, the next largest time sink is discussed by referencing a UC Berkeley study which counted time validating fares, boarding, alighting and bike loading. No data was reported as to bus design effects on dwell.

    In recent years, AC has purchased buses with very slow operating doors. In these units the driver enables and then passengers activate a door with a slow mechanism. There is a driver option to immediately open these doors at stops, but AC’s driver manual explicitly outlaws this procedure. Some riders choose the front door as the faster way to alight thus exacerbating dwell. . In turn, the time to load a wheelchair increases with the driver having to operate a console away from the driving seat. For reference, SF Muni specs double stream doors in order to decrease dwell.

    As the study makes clear, the current use of ‘dip swipe’ fare collection is even slower than fumbling for change. AC’s decision to cease honoring BART Plus style ‘flash passes’ has increased time wasted in fare control for the average rider. The study notes that UC students and staff DO use a flash pass and board rapidly. While Translink is nearly as fast as a flash pass, market penetration is miniscule so far.

    Several route splitting, express/local, and A/B route changes are discussed.
    First A/B skip stop works best when discreet travel destination/origin pairs exist because of either concentration at one end or major neighbohood differentials in transit needs—none of this characterises the 51. While ideally an express or limited stop service in base day service should be useful, IF likely origin /destination pairs are well researeched, as 80% of delay time is signal caused, expresses will be equally hampered. Thus the express will have little time advantage over the current route. Any of these route splits, however, bring up a systemic issue—transfer rights. When I moved here in 1970 and for many years therafter, a transfer was free and good for TWO uses. Now the single use restriction acts to constrain travel patterns, and clearly splitting the 51 in any way will neccessitate a change in policy because many riders ‘cross’ the potential route boundaries.


    Scrap dip swipe for more contactless media—flash or tag.

    Equip entire fleet w/signal preempts—this will speed ALL routes.

    Work w/ City of Oakland to open up the bottleneck Alcatraz to Claremont

    curtail stop eliminations—if the ridership is as low as stated, MOST runs never stop there now. No cost to retain for the convenience of riders

  5. jennconspiracy March 25, 2009 at 9:17 am #

    Not only do we need a rapid 51, it should just NOT stop from the turn off of Shattuck until it gets to, say Claremont. If we could have a 51 that doesn’t stop anywhere in the university area, that would help folks who really need to get from University Ave to Broadway. Seriously – the college kids who get on, travel 2 blocks, and get off – need a slap on the back of the head! They slow down the bus and it’s really a PITA – they get to use the bus for free so use it needlessly. You’re young – WALK!

    I have often walked from University/Bonar to 51st/Broadway rather than take the 51. I even got off the 51 when it took over 50 minutes to get to College (it was so crowded!) and walked home from there. It’s one of the most useless buses for distance travel.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: