A map of the area around Eastgate Park and Ride where Ride2 service will work. Eastgate Park and Ride service area

King County news release

King County Metro customers will soon be able to use new mobile apps to hail an on-demand shuttle to and from transit hubs throughout the region, starting at the county’s largest park-and-ride.

Starting on Oct. 23, commuters will be able to use the first app – called Ride2 Park & Ride – to hail a shuttle operated by Chariot and Ford Smart Mobility to and from the Eastgate Park-and-Ride in Bellevue.

Metro will make similar on-demand shuttle service available to other transit hubs throughout King County over the next several months as a part of a yearlong pilot project.

The pilot will be free for a few months, at which point you’ll be able to board using your ORCA card.

Much like the trailhead service, which started in a limited fashion in 2017 and then expanded in 2018, Metro is going to take advantage of the Chariot partnership to try this and see what works.  Metro’s Jeff Switzer told me that Chariot was selected through a competitive bid, and that the two organizations are splitting the costs (though Chariot picked up the tab for the app development).

After the two month pilot, Switzer says, “Metro will begin charging standard fares, and we and our partners will analyze the ridership data and cost data to determine the best pricing model for continuing the service and possibly expand.”

One potential stumbling block is Chariot’s use of nonunion drivers.  KUOW’s Joshua McNichols noted that the ATU was “nervous” about the service and might make a unionization push.  Union opposition helped kill a last-mile service in Seattle earlier this year.

All over Puget Sound, agencies are experimenting with different flavors of “microtransit.”  Mercer Island is using some Sound Transit settlement money for it, Seattle Mayor Durkan’s latest budget proposes a last-mile pilot, while Pierce Transit is running subsidized Lyft rides to and from transit stations.

Nationally, results are mixed. Streetsblog wrote critically of microtransit experiments, but concluded that the best of them were performing “like decent dial-a-ride services.”  Given that the Eastgate service is peak-only and serves a limited geographic range, it seems unlikely that it would cost as much as dial-a-ride. But that’s what the pilot’s for, I suppose.

Assuming the subsidy levels are manageable and the union question can be resolved, a service like this is probably the right approach for the Eastgate area, which has a full park-and-ride and the low-density, hilly, land use patterns that are the bane of effective fixed-route transit service.

21 Replies to “Metro and Chariot Partner on Eastgate Microtransit Service”

  1. “Union opposition helped kill a last-mile service in Seattle earlier this year.”

    Give the firm some credit for valuing its good name, Frank. Because it gets hard-earned credit for realizing its hiring and maintenance budgets can’t anywhere near afford the expense and exposure that service of this kind invariably carries.

    And I’m sure they’re grateful that the union was there to help them with this painful but necessary decision. Thanks, Frank, for making sure that both these entities get credit for their conscientious attention to the reputation of passenger transit.

    Mark Dublin

    Mark Dublin

  2. This seems like a weird one-off solution similar to dial-a-ride transit (DART) routes. I hope they are evaluating this as a way to modernize and replace DART routes, whereby they keep a fixed route like they do today, but the app is used for reserving off-route pickups and drop-offs. I would suspect that they would have to change contractors from HopeLink to make this work, but I think it would make this service much more utilized and practical.

    1. DART routes exist to provide limited off route deviations on a fixed route while this new service is exclusively designed to be the first or last mile. Most DART users use the fixed routing, but every DART route (by definition) has an area riders can call ahead and ask for a deviation.

  3. Couple things might work better, Frank. One, contract with one or more cooperatives- eliminating need to pay profits. Used to be very common for cabs and buses. Problem being that if you’e part – owner, you’re also the whole boss, meaning brutal hours and no union representation.

    But better one: just accept that The Last Mile (sounds like a war movie, doesn’t it? It really is one, isn’t it?) is part of the route map for one route or more. Without which the line haul vehicles would carry no ridership, and the would-be passengers’ cars would block the main routes’ lanes.

    And give drivers whatever average Uber driver wears. Pin-striped suit? Hoodie? Fact I can’t use a touch screen means that for me it’s all hearsay.

    Hey, Instead of betting what’ll work, let’s put it into the State Lottery on condition we don’t wrap any windows with it.

    Drivers Metro and ST- hired and trained- assuring max efficiency. Or at least removing any excuse for its absence. Maybe I’m thinking of “A Bridge Too Far”. Meaning the one where I-90 ramps into Bellevue way. Or is it about Spokane Street early next year?


  4. Jarrett Walker has some excellent articles on microtransit. Microtransit does scale well because it inherently consumes too much labor per passenger pickup to carry a non-trivial number of passengers for acceptable costs. This is because every passenger pickup requires at least several minutes of driving specifically to get to/from that particular passenger’s destination. Since all the requests are on-demand, it is not possible to coordinate departure times in advance, so that people whose pick-up points are all “on the way” can actually get picked-up in the same vehicle, traveling a straight-line path.

    Even when passengers traveling at similar times are on the way, you’re going to case where a ride request comes in right after the van has just passed the passenger’s pick-up point, so what would have been a very quick stop with advanced scheduling now requires time-consuming turnarounds. I had this happen to my on UberPool not that long ago. We were going northbound on Montlake, coming off the 520 bridge and seconds *after* passing the UW Link station, a ride request came in from a passenger who wanted to be picked up there. The turnaround easily ate up a good 5-10 minutes. While this microtransit pilot is certainly great for coverage (e.g. providing service to a handful of people who live in that area, who would otherwise not have bus service), I don’t see this having an impact on Eastgate-area bus ridership beyond statistical noise.

    If the sole problem is really *just* a matter of parking filling up, and you’re willing to make the assumption that every single person owns a car, you could probably get more new riders for the money by simply hiring valet attendants to park cars in the aisles, when the regular parking spaces become full. But, of course, such an approach completely fails those who commute in the reverse direction, or don’t have cars, people whom the microtransit approach *does* serve.

    Overall, I do think the microtransit pilot *is* worth trying, but it’s important to go into it with the right expectations, and understand that this is a *coverage* service, whose goal is to increase the number of people who *can* be served, not a ridership service, whose goal is to actually carry a lot of people. Best-case, the microtransit pilot can reveal enough demand on corridors Metro never thought about, to justify new service on a fixed-route bus.

    1. Sounds like a network delay situation to me. Takes a few seconds from the time a rider requests to the time the driver gets the instructions to alter the route. UberPOOL programmers should know better than to “trust the network” for anything faster than a minute or so.

  5. asdf2, are you saying that with passengers already aboard, your driver took ten minutes turning around and going back to pick up someone else? Giving me to understand that there’s no minute to minute coordination at all?

    When I saw these services begin in the late seventies, they were always referred to as “paratransit”. And either passengers had to call hours, or even a day, in advance. Essentially setting up the routes so this kind of thing would never happen.

    Or, computer dispatch setting up routes and pickups while the fleet was moving, so passenger that one van just missed would be first on the next vehicle’s schedule.

    Let’s just admit that what’s needed here is taxi service. And that most reliably way to run it is that transit owns, repairs, and trains it. Would still mean clearer roads than if every passenger used their own car. Though with two and a half years’ cab driving on my resume, either contract with Orange Cab or steal their drivers.

    But the real sweat-level work, which I’m still waiting for somebody to address, is how to change existing living patterns so that both transit and life work. In the aftermath of 2008, we lost a generation’s best chance, when thousands of foreclosed and unsalable houses could’ve been painlessly bought, leveled, and their space Transit-Oriented.

    Call it #D-Sprawl. Starting with whatever rules and laws prevent transit from getting into residential development, That’s what shredders and dumpsters are for. Though given unmistakable trends upwind of us, good idea to have next move on the boards ready to go.

    Mark Dublin

    1. In this case, I was the only passenger on board, so the detour isn’t nearly as serious as when a bus holds up 50 people to wait at stoplights to go in and out of a transit center.

      Nevertheless, this was perfect example of how UberPool is nothing like real carpooling. My trip was from Kirkland to my home near the U-Village. We drove all the way to Seattle, with just me and the driver. Then, immediately after passing the UW Link station, we got a ride request of someone who wanted to go one mile into the U-Village. So, we spent 10 minutes turning around and waiting for that interminable left turn light from Montlake into the U-Link parking lot. Now, we actually had something resembling a carpool for the last mile or so.

      Except, again, not really. I live west of 25th, and the other passenger was headed to the shopping center itself, east of 25th. So, after about 1 minute (3/4 mile) of both passengers actually riding the same direction, the way a carpool is supposed to work, Uber’s algorithm now has to choose which passenger is going to be dropped first. Whichever choice it makes, the car is guaranteed to spend more time driving in circles, traversing pedestrian-scale distances than actually carrying both me and the other passenger for the segment of the drive that our trips had in common. And that’s not even getting to the fact that for 90% of the distance for my ride, I was the only passenger.

      One big difference is that when people arrange carpools among friends, everyone is courteous and respectful of everyone else’s time. But, when the fellow passengers are complete strangers, the thought process is different. The other passenger’s time no longer matters to you and, instead, you feel like if you’ve suffered the inconvenience of driving around in circles to pick them up, but don’t get door-to-door service to *your* home in return, you haven’t gotten your money’s worth.

      In theory, Express Pool was supposed to mitigate this issue by potentially asking people to walk a short distance, in order to make the routing more efficient, but in reality, Uber is pessimistic enough about one’s walking ability that it ends up not really making any difference. I once rode it to Carkeek Park, taking another passenger to Northwest Hospital along the way, and we *still* drove all the way around through the hospital parking lot to the building’s front door, instead of just dropping the passenger off at entrance. For reasons mentioned previously, if this were a ride among friends, the passenger would likely have volunteered to walk from the entrance in order to save time for everyone else. But, with Uber, you have zero incentive to do that – since you’re paying the same amount anyway, so you may as well get your full money’s worth, the time of the driver and other passenger, be damned.

      At the same time, what Express Pool *doesn’t* do is tell you which side of the street to wait on, so that the driver can pick you up without having to turn around or wait for you to cross.

      1. “What side of the street” is a constant issue with ride hailing apps. Especially in situations where it’s neither legal to u-turn nor to cross the street! There’s no way for riders to even specify what side of the street they are on, and GPS is far from reliable enough. The addresses driver’s see are tiny print and require switching from the GPS screen, at least on Lyft. And that’s assuming the driver knows which side of the street odd and even addresses are on that particular street. Just put a feature in app for riders to specify which side of the street–I could probably code this in a week!

  6. I am probably more hopeful than most about this pilot. Given that Eastgate P&R fills up pretty quickly most days, I’d rather see Metro spend money on this pilot rather than building more parking. I think they’ll find a fair amount of demand during peak from the business “parks” (what a funny term) nestled in the northeast corner of this area, and all day from Bellevue college students & staff. The 554 is all day (not quite at the P&R but short walk up stairs and over the bridge) and people would find that more useful if they could more easily get to it.

    This certainly beats 30 min neighborhood “circulators”. I wonder at what level of usage they will deem this a success and keep it going?

    1. At least for Bellevue College, most people will simply walk, and those that don’t want to walk have plenty of fixed-route buses to choose from, which have a very high combined frequency, during the peak hours. Fixed route buses are far more efficient than microtransit vans, and if people end up switch to the microtransit, just because the van happens to be already at the P&R, waiting for passengers while the regular bus is 5 minutes away, that’s making the whole system less efficient, and taking the van time away from people going to places that really need it.

      The business “parks” in the area are generally already served by the 271 and/or private employer shuttles.

      I don’t see this making any dent at all in Eastgate P&R capacity and, if parking mitigation were really the end goal of this, valet parking in the aisles would probably do it more efficiently than microtransit rides between home and P&R (at the expense of abandoning all pretense of serving people without cars, not connected to fixed-route buses).

      1. The write-up doesn’t make it clear whether the pilot is intended for first-mile usage (from residences to the transit center) or last-mile usage (from the transit center to offices), although some hints lean towards first-mile.

        What is great about this service area, and could make the pilot a success, is that the service area includes lots of homes and jobs. A van could pick up a handful of people in neighborhoods, bring them to the transit center, and then fill up again with people leaving the transit center for nearby office parks.

        I don’t see much midday/evening/weekend demand, however, which reduces the overall cost efficiency, unless the vans and drivers find some other profitable activity during that time.

      2. @Chad — Care to address the other issues discussed below (https://seattletransitblog.wpcomstaging.com/2018/10/20/metro-and-chariot-partner-on-eastgate-microtransit-service/#comment-809586 along with the responses)? Assuming this is popular, what does that mean for the 212? Don’t we need to add more runs? If so, wouldn’t we be better off just sending a similar bus to a nearby neighborhood (or extend half the runs to a nearby neighborhood) instead of running more 212 buses *AND* this expensive point to point service?

        People aren’t really trying to get to Eastlake (for that there is actually pretty decent bus service). They are trying to get to downtown Seattle, and Metro has basically said the only decent way to do that is to drive to Eastlake and take the bus. Micro-transit to Eastlake misses the point — you just need better bus service to downtown Seattle.

  7. Microtransit seems like a solution looking for a problem. We already have DART, and this seems like the wrong tool for that service. It is reminiscent of charter schools. While billed as something revolutionary and new, it is something that has existed for years. The only significant difference is that it manages to bypass the union, with the unfounded assumption that higher paid union workers are less effective than union members.

    If you really want a capitalist answer to this problem, do the obvious: charge for parking. Then put the money into other bus routes, that serve other park and ride lots. The lots should be closer to the neighborhoods, and the buses shouldn’t be shuttles, but through routes. It isn’t a lot different than what happened with the 41. Initially, park and ride users drove ridership. The main lot was huge, but they had a second one a few blocks away. That was replaced with a park, while development occurred in the neighborhood. While the main lot still exists, and I’m sure parking there is difficult, there are small satellite park and ride lots (https://goo.gl/maps/3twvBGLPhhy). That means that there is little reason for someone in that part of town to drive to the main park and ride. As a result, no one is asking for a bigger Northgate park and ride lot — it has been the opposite — community members pushed for a smaller lot.

    If there is opposition to charging for parking, then go ahead and waive the fee for those that ride the bus. This would mean that immediately you have data on who is using the system. Someone uses ORCA for the parking, then ORCA on the bus (or a token for both). With the data you can map out a better solution (better bus routes and good satellite park and ride lots). That is probably the one decent argument for this service — they are gathering data. But ultimately, I don’t see this as a cost effective long term solution.

    1. The more I look at this, the more I understand why Eastgate Park and Ride is so full. For many, there is no alternative to getting downtown. If you live within the service area in that map, you don’t have much choice. If you are lucky, you might live close to one of the corridors with connecting bus service, but even then, it is not necessarily frequent, nor is it the direction you are headed. So that means a poor two seat ride (and only for a handful). If you drive to one of the other park and rides, then you have a two seat ride. All the buses that serve those other park and rides (Newport Covenant Church, Eastgate Congregational Church, etc.) have buses that go to Bellevue, not downtown Seattle. The 212 is a fairly straight shot right into downtown Seattle. It runs every six minutes in peak direction. If you have to leave work early for some reason, you can take the less frequent, but still reasonable 554. That is a great commute, if you can get a parking spot.

      The 212 simply ends at the park and ride. One possibility would be to extend some of the runs to Saint Andrews Lutheran Church. They have a pretty big lot, and my guess is you could negotiate with the LDS temple and get some of their lot. Such a bus route would also pick up some of the apartments close to the college, which would otherwise be a hefty walk to the bus (https://goo.gl/maps/DweNmpvFMRN2). Even if it just looped around the Shell station (https://goo.gl/maps/HQ71nSsYZnN2) you would have easy overflow from the park, which means that if the park and ride is full, folks park at Robinswood, and walk five minutes to their bus. With service every six or seven minutes from Eastgate to downtown Seattle, doing that sort of thing would be easy, and very similar to what Metro did with the 41. About half the runs head out there (giving folks 12 to 14 minute service) while the other half just end at Eastgate.

      Another option would be to run an express serving Newport Covenant Church. They appear to have a very big lot, and room for more. Again, a little negotiation, and they can pave part of their lot, adding a lot of new spots, which is great for everyone. Like the extended 212, the bus route would basically just turn around by the park and ride. As it turns out, there is precedent for this, as the 245 does this. This new route would therefore look something like this: https://goo.gl/maps/Xc2VUFCkkmS2. Like the extended 212, it would mean connecting apartments along the way (in this case, the apartments off of Factoria) directly with downtown Seattle. That sort of bus works in both directions, as a means to serve Factoria businesses (making for a more efficient bus). Serving Factoria in that manner means that the little side trip of the 212 to serve 36th becomes less important. That could lead to the 212 directly serving Eastgate, saving some time from a service standpoint (to make that happen, you would probably need to pair the 217 with this new bus route or the new bus route would have to be frequent enough to be a good alternative to the 217).

      None of this is free, but it isn’t that expensive, either. It would lead to a major improvement in service for a fair number of people. The 212 could be even more an express. Some of its runs would extend to serve people in the neighborhood (who live in the apartments to the north). The other bus route would provide people who live south of the I-90 and west of 520 a fast alternative to the options today.

      1. When RapidRide B was created, Metro reshuffled its bus service so that nearly all Eastside->Seattle service (the 255 being notable exception) was consolidated into major P&R hubs. Before the restructure, most residential areas of Bellevue actually had direct service to Seattle, at least during peak hours.

        The catch was that the peak-hour service to the neighborhoods ran only every 30 minutes, while off-peak service was limited to a half-hourly or hourly milk run at best, if the bus ran at all. In many cases, the peak express to Seattle would the a neigborhood’s only bus route, which meant people simply had no transit service outside of rush hour, and even during rush hour, people trying to get home from places other than Seattle had to wait for a half-hourly bus whose punctuality was at the mercy of traffic on the I-90/520 bridges and within downtown Seattle, itself.

        The restructure traded away peak-hour direct-to-Seattle service to all the neighborhoods, in exchange for very frequent and more evenly spaced buses at the P&R hubs, plus all-day, 7-day-a-week local routes to neighborhoods that previously had no service at all, except their rush hour express to Seattle.

        While the restructure greatly benefited people who drive to a major P&R, or make local bus trips within Bellevue, it had the side effect of making that drive to the P&R a necessity for many more people. Theoretically, the feeder buses existed. But, they are infrequent, and, in many cases, indirect, subjecting riders to a grand tour to Bellevue College simply to connect to the bus to get them to work. On top of that, making the bus connection without a long wait is quite difficult. During the AM rush, it’s not too bad, simply because the Eastgate->Seattle buses run so often. But, getting home during the PM rush, unpredictable traffic means you can’t just choose a Seattle->Eastgate bus to make the schedules line up, without risking a minor traffic delay cascading into a 30-minute wait. And, during off-peak hours, yes, the feeder buses are running, but the schedules are mis-coordinated so that a midday transfer from the 554 to a route like the 221/226/245 requires far more waiting than it should. And the fact that the 245 drives right by the Eastgate freeway station without stopping there doesn’t help matters either.

        I’m not sure Metro anticipated when they did the restructure, Eastgate P&R filling up, like it did, but, in hindsight, it’s not at all surprising.

        As to the solution, I suspect that it would be possible to extend many of the route 212 trips to at least some of the neighborhoods at very little cost, since the buses are already deadheading from East Base to Eastgate P&R in the morning, and from Eastgate P&R back to East Base in the afternoon. With a little thought, it seems these buses ought to be able to serve some of the neighborhoods on the way to/from the base without having much impact on runtime and operating cost.

      2. @asdf2 — That makes sense. That is why this is really a good problem to have. It doesn’t signify a failure, only that one aspect of the change (using the Eastgate Park and Ride to take the 212) is more popular than they planned. The answer is not to expand the park and ride, or look at this as something that requires a DART style system to fix, but simply to expand on the same model. Send buses to other park and rides and people will follow. When the parking lot is full (and very big) and the buses run every six minutes, then it is time to steal away some of the riders with other routes. Ideally you run the bus along a corridor that also picks up walk-up riders (as both of my suggestions would) but if the bulk of the customers are park and ride users, so be it. I think either suggestion would work at very little additional cost, because the 212 must be fairly full. You don’t run buses to downtown every six minutes because you want a better experience for your riders — you run it that often because the bus is full. That means that if this experiment really does work, then you are running more 212 buses anyway, and these buses could serve other nearby neighborhoods at roughly the same cost (thus eliminating the need for the microtransit).

        I really feel like this aspect wasn’t considered. Assume this works out great, and a thousand people each day use it. Doesn’t that mean you need to run more buses? Where does that money come from? It is as if folks didn’t really look at the big picture here. Folks aren’t really trying to get to Eastgate — they are trying to get to downtown Seattle. I really doubt there is unused capacity with the 212 (unlike, say, Sounder or Link). Running the 212 more often does very little to improve the system, and costs just as much as running a bus to a similar area (e. g. Factoria).

        The more I think about it (and the more I learn about this — thanks asdf2) the more I think that this is just an excuse to try out a new, union-free form of micro-transit.

    2. By the way, I would feel differently if the 212 wasn’t so frequent. If the 212 ran every fifteen minutes, then there would be a case for concentrating service there. But right now it runs very frequently.

      Just imagine if this is successful. Imagine that this results in a lot more people getting to Eastlake, to take the bus. What then? Well, you have to run more buses. How about they run 16 buses an hour, instead of 10. That is nice, but it really doesn’t improve things substantially. It means buses around every four minutes instead of every six. Big deal.

      In contrast, running six buses an hour somewhere else makes a huge difference. That means ten minute service to a different neighborhood. For many of those riders, it involves no shuttle — just a simple walk. For others, it means driving to a different park and ride, but one that is closer and not right by the freeway. This shuttle thing idea is nice, but it is hard to see it actually leading to a much better system, despite the extra cost.

  8. For some reason I thought this article was going to be about Cherriots, the transit operator in Salem, Oregon.

  9. asdf2 overstates the fall 2011 changes. one-way peak-only routes 225, 229, 261, 266, and 272 were deleted; those deletions were well-mitigated. one-way peak-only routes 250, 260, 265, and 268 were retained; later weaker routes 250, 260, and 265 were deleted in fall 2014 with the reductions. the fall 2011 network connects Eastgate with routes 221, 226, 245, and 271 to the north of I-90, as well as routes 271, 241, 246, 240, and 245 to the south and east. Let’s watch to see if the new service is cost-effective and what happens to overall network productivity. It is a trial.

Comments are closed.