Yesterday at the Transbay Joint Powers Board meeting, Ben Tripousis of the High Speed Rail Authority proposed an alternative approach to Caltrain/High Speed Rail platform compatibility that could leave more time to assess the most effective solution for the blended system.
At the meeting, Transbay board members heard presentations from Caltrain and High Speed Rail about platform compatibility, and were very attentive to the compatibility issues that could affect the amount of passengers who can use transbay in the future. Board member and Muni CEO Ed Reiskin commented that “We are building this system for 100 or more years. Compatibility will be essential for future capacity. The board wants to review additional information and provide policy input to the decision to be made by Caltrain and High Speed Rail.
In his presentations, Tripousis proposed an idea to keep options open while High Speed Rail and Caltrain work out more information about how the blended system will work over time.
Currently, Caltrain is considering providing level boarding at a height compatible with High Speed Rail’s high platforms. The proposed approach is to buy cars two sets of doors – a low set of doors at 25â€, and a high set of doors at ~50â€. Cars would be configured with fewer seats order to clear the way for the extra doors.
Dual doors for Caltrain/HSR compatibility
Caltrain would upgrade platforms to the higher High-Speed-Rail compatible height over time. Caltrain could upgrade all of the platforms to 50â€, and then the extra doors would be sealed up and the seats would be add back. Or Caltrain could upgrade only the platforms at High Speed Rail stations, at Transbay, Millbrae, Diridon, and maybe Redwood City, and then the cars would have fewer seats always. (Note that in order to get to level boarding and longer platforms and longer trains to carry more riders, Caltrain will need to upgrade all of its platforms over time anyway)
There are a lot of un-answered questions about the best path for Caltrain and the blended system. Answers to some of the questions will be figured out when High Speed Rail resumes its planning for Peninsula service and the blended system starting this summer.
Would compatible platforms enable the High Speed Rail authority to modify its designs for the Millbrae and Diridon stations to be more compact, less expensive, and providing a better rider experience?
How would platform compatibility affect overall ridership of the eventual blended service – with regard to blended commute service, where a commute passenger between SF and SJ would be able to choose any of 10 trains per hour; and with regard to Caltrain as a feeder system to/from High Speed Rail
The different scenarios have different impacts on dwell time – the amount of time the train is at the station letting out and picking up passengers. How will dwell time affect the reliability and schedule quality of the blended system?
Tripousis described an option whereby Caltrain and the High Speed Rail Authority keep options open while these questions are answered. Caltrain could order cars that had working doors at 25â€, and “ghost doors†at 50â€. These would be slots in the car design, that could be opened at a later date to create doors at 50†boarding height. Depending on how and when Caltrain and High Speed Rail decide on a compatibility solution, the upper doors could be opened, where needed, and the lower doors could be closed, eliminating or minimizing the amount of time that Caltrain would need to operate with fewer seats.
What do you think about this option for keeping options open?
The problem to avoid is rebuilding Caltrain platforms twice.
If Caltrain orders EMUs with doors at 25″ and inoperable door cut-outs at 50″ … we don’t get level boarding until platforms are rebuilt to at least 25″. Once HSR comes online at 50″ and Caltrain converts the 50″ cut-outs to real doors, we lose seats until the lower doors can be discontinued … which will require rebuilding platforms which were already rebuilt to 25″ to 50″.
Rebuilding platforms from 8″ to 25″ or from 8″ to 50″ costs about the same, so why not just start with real doors at 50″ so Caltrain stations are rebuilt to that height the first time to achieve level boarding?
If they go with the phantom 50″ doors we can’t have level boarding for a very long time or, if we do get it at the 25″ doors then most stations will likely remain stuck at 25″ for a long time … or they will need to be rebuilt twice in order to reach a systemwide single shared 50″ platform height where any train can call at any platform and for Caltrain to be able to plug the 25″ doors and regain the seats lost when the 50″ doors were activated.
Two sets of doors is probably the only reasonable solution I’ve heard given the situation.
The idea of keeping options open is probably the most suitable path Caltrain can take without making any kind of commitment. Caltrain cannot commitment to a 50″ platform height because it doesn’t work well in all other circumstances (except sharing platforms with HSR) and I don’t think Peninsula communities support it.
In the future, there may be circumstances where Caltrain communities would want to go high platform, or that there will be a solution at Transbay to allow trains at two heights to share platforms.
I don’t see a case for Caltrain to build 50″ platform everywhere else. It slows down bike boarding times. In 25″ platforms are built everywhere else and 50″ at Transbay, then the only time that high door will be used is at Transbay, at a time where dwell time is not an issue and where cyclists can hold until everyone else gets off the train. I can forsee bikes moving bikes up and down will be a nightmare at stations like Palo Alto where turnovers take place.
50″ doors don’t slow bike boarding any more than climbing the 3-4 steps today in Gallery cars with only a single door. In fact, it will be about same speed as current Bombardier bike boarding. Assuming that bikes will stay on the lower level, climbing the 3 steps to the mid-level 50′ doors won’t be bad since there’ll be some room to stage the bikes by the doors.
Yes, but the goal of level boarding is supposed to be much better dwell time and reliability. The goal isn’t to avoid getting worse!
– Video: http://sanfrancisco.granicus.com/MediaPlayer.php?view_id=29&clip_id=22742 (click on #15)
– Transcript: http://sanfrancisco.granicus.com/TranscriptViewer.php?view_id=29&clip_id=22742
Let’s keep options open. I like this.
“Ghost” doors don’t relieve us from short-term decisions about future platform interfaces. One of the first things you do when you engineer a complex system is to define and control the interfaces. The HSR people, Caltrain, and Metrolink or whoever is the SoCal stakeholder need to hammer out NOW where the platform edges will be. This is not a decision that can be delayed any more.
The problem with interior steps for bikes is that people will try to move bikes up and down the steps when trains are moving. While today people have to haul bikes up and down, but they only do that while the train is stopped.
If TJPA insists on high platform for Transbay, then that might be the outcome since the injury risk of bikes moving on the steps are much lower if it is done only at Transbay and where Caltrain and effectively stop people from moving bikes up and down the steps while the train is in motion.
Why should Palo Alto, or any other non-HSR stops, use 50″ platforms? It is unsafe, it is unsightly, it is costly. It may make TJPA or CHSRA’s job easier, but local communities also have a stake especially when it comes to funding.
Marian Lee is correct about TJPA making positions that affects other agencies. If SF politicians aren’t ready to commit a 100% ADA access for Muni Metro at all stops, SF politicians have no right to tell other cities how high platforms they should use. How would Ed Reiskin and city supervisors feel if MTC or other bodies telling Muni to put high platforms at all street level Muni Metro stops? Who could disagree with 100% ADA access and dwell time reduction for Muni, right?
The TJPA has every right to take a policy position on issues that will determine the success of the Transbay Transit Center, regardless of how popular those positions may or may not be elsewhere on the peninsula. Not taking a position because it might be controversial is silly.
The Caltrain bike issue can be figured out; it is by no means a fatal flaw. The danger is in lifting a bicycle while navigating steps on a moving train; however, there is no need to lift one’s bike if said steps have suitable wheel gutters. Two feet and two bicycle wheels with brakes applied make for a very stable posture, easily held and released even while standing on the interior steps of a moving train. As a Caltrain bike rider, I feel this issue is being blown out of proportion.
Bikes on board shouldn’t drive the architecture of the entire state-wide rail system, which is what is at stake here. It is an important issue, but it needs to take its place among other issues that are even more important, such as the cost/benefit ratio of the massive infrastructure investment required to build a workable blended system. Blending is a package deal, for better and for worse for both operators, and until four-tracks-everywhere do us part. (Ha!) Caltrain can’t just take the HSR money and run.
It’s a good, reasonable option — ‘within the box’.
Another option would be to run ultralight, composite vehicles in multiple short trains, some with high doors and some with low, or to have some cars in a train have high, and some low, doors, in a consistent way that allows for facilities at platforms. If cars are shorter they can fill faster. It’s time for a total re-design of railcars, high- and low-speed.
@Mark Roest. This is precisely the reason why Bombardier decided to remove all doors and toilets from the Omneo passenger cars and moved them to the short (33 feet) motor cars which can be ordered at any floor height (up to 920mm for the UK), so it would be possible to alternate high and low floor motor cars if needed. Having said all that, there is a fairly good chance that the RSAC ETF will figure out a solution to this dual-height platform nonsense before we order the first EMU.
@Clem: Correct: This is not a decision that can be delayed any more but, as we all learned the hard way, the fine SamTrans and HSR teams did not consult with the RSAC ETF whose job it is to fix the mess they created: http://www.apta.com/mc/rail/previous/2011/Presentations/R-Lauby-Development-Of-High-Speed-Rail-Equipment-Interoperability-Requirements.pdf
@ Clem: nobody disputes that the TTC needs compatible platforms for HSR, Caltrain, Capital Corridor and anything else that might ever go there but platform heights are the least of the TTC train box problems. First, the platforms need to be shifted west about 200 feet to make room for the tail tracks and, second (as Director Ayerdi-Kaplan mentioned at the end of the video), the issue of compatible widths also needs to be addressed.
@Andy: Every additional pair of doors eliminates 16 seats (8 on each side) and every set of stairs eliminates a further 8 seats (or 8 bikes). This is the reason why the front stairs were removed and the floors were raised in the cab cars in the baseline configuration (this is what made it possible to add 8 bikes (total 40) in each cab car). http://www.greencaltrain.com/2015/04/might-caltrain-trains-have-less-room-after-electrification/
Hilarious how Caltrain buried this rather important point in a footnote:
*Double ~50†doors may not be feasible
There goes your dwell time!
The agencies involved are trying to find a solution that is workable for all parties. One party making a positional statement can jeopardize the process. If somehow SF thinks it can bully Peninsula cities by passing resolutions, the cities can in turn do the same to deny HSR to SF.
The bike issue needs to be front and center during the design. We are not talking about accommodating something that has yet to exist, but accommodating something that is already out there. Bikes are not going away, and the bike community isn’t going to accept the notion that somehow onboard space will be substituted by at-station bike storage to address the bike problem. They only want more of it. Frankly I don’t know how they’re concerned about dwell time and overall trip time.
Also I don’t think the in-vehicle wheelchair lift is going to be ADA compliant either. The whole common boarding height issue is driven by what it seems to be an ADA mandate for level boarding. While from the specification perspective the two door solution meets ADA, but yet from the experience perspective it is not an improvement.
If we make the Transbay platforms all 25″ tall, and use portable ramps for all HSR doors, may be that would address the compatibility issue and offer ADA goal of equal and non-assisted boarding by disabled passengers for any train.
Caltrain isn’t about to dictate anything about platform interfaces for HSR. The tail doesn’t wag the dog. Bikes on board aren’t about to dictate the entire architecture of California’s blended passenger rail system, starting with the EMU procurement. The tail doesn’t wag the dog. Peninsula cities worried about high platforms impacting their historic depot aren’t about to tell San Francisco how to configure their rail terminal. The tail doesn’t wag the dog. These aren’t opinions: they are simple observations…
No need two different height of door – it is very expensive EMU!
Transbay terminal will have 3 platform and 6 tracks. 4 will be given to HSR and 2 for Caltrain.- Looks enough. Taipei, Tokyo (JR east) both are handles lots of HSR with 4 tracks.
For Caltrain 2 tracks should be good enough. As we know, Fremont, Richmond BART handles 8 regular train plus some other deadhead in every hour. Why electrified Caltrain need more infrastructure?
Some of BART train are coupled or decoupled at Fremont station before or after peak period, which takes only 1~2 min of time.
Clem, what’s the tail and what’s the dog?
By, say, 2030, how many passengers o you estimate will arrive at TBT per week on Caltrain, as opposed to CAHSR? Which passengers should have priority from the perspective of TJPA, and why? What are the costs of equipping HSR with retracting steps that will accommodate use at 50″ and 25″ (or 550, 750 mm) platforms, as opposed to placing the onus on Caltrain (and perhaps Metrolink)? ADA does not require full compliance at shared-use stations.
@Kim: none of your examples include the additional challenge of conflicting all inbound traffic with all the outbound traffic of another operator, in order to reach the single assigned platform.
@Marc: Caltrain passengers will always far outnumber HSR passengers, which is why all Caltrains must serve Transbay (despite Caltrain’s objections). While it would be nice to see the TTC designed around Caltrain’s requirements, you don’t have to be very astute to notice that HSR is calling all the shots. It is the de facto dog.
You should know that California does not have the “get along with the program” attitude towards pretty much everything, and tend to be a trend setter. California has a tougher auto emission and gasoline standards. If California gets with the program our gas prices would be cheaper. California’s auto emission standard will be enacted nationwide for the next model year.
California is the first to enact smoking ban, and other states have followed. California is the first to attempt gay marriage, and despite all the set backs, it is legal and many other states followed. California is the first state in the nation to legalize medical marijuana.
So the idea that bikes onboard are less important than California adopting a Northeast Corridor standard isn’t going fly. And I think Californians are less concerned about “proven standards” if following those standards would leave certain needs unaddressed.
Caltrain can be argued as a trend setter for allowing bikes onboard as well as pursuing non-compliant EMUs for electrification (other systems including SMART and Denver RTD opted for FRA compliant vehicles.) We pursue that because it best meets our operating environment, and that we are willing to take risk doing it. Shouldn’t Caltrain have placed to buy EMUs off RTD’s or SEPTA’s options? Considering that RTD will use NEC platform heights?
If California is so visionary and forward thinking, then why are we not building our HSR system with 30″ floors, an approach that would satisfy the needs of both operators with the least amount of disruption? Where has the rail advocacy been on this issue and why is there a total failure to think Californian? Why are you not pushing for a compromise solution?
Instead of pushing HSRA to choose 30″ height, all you’ve focus on is how Caltrain should adopt 50″ height, even with all the operational and transition issues involved.
Dual boarding heights with eventual conversion to 50″ platforms is the best available option that remains if HSR refuses to consider 30″ platforms. It still achieves the underlying goal of interoperability. I am not “pushing” this idea: you don’t see me attending every board meeting or whispering into the ears of Caltrain staff. It’s their idea, a reasonably good one, and I hope that it goes somewhere.
Meanwhile, you are advocating for Caltrain and HSR to go their separate and incompatible ways. Take the HSR money and run, and forget compatibility for generations to come! This is a position that I believe is far more harmful to Caltrain’s interests than dual boarding heights.
As for transition issues, they are almost trivial with dual boarding heights. The logistics of the transition from 8″ to 25″, on the other hand, are quite murky and undefined. It is not at all clear how EMUs with boarding on the lower level could support an evolving mix of 8″ and 25″ platforms without imposing significant constraints on operations, platform construction phasing, and even vehicle design. Caltrain’s “original plan” for level boarding at 25″ wasn’t a plan. It was a vague idea that was sketched out on the back of a napkin when a few stakeholders complained that the level boarding issue wasn’t being considered.
I am glad they are now considering the issues and challenges that come with transitioning to level boarding. There is no “easy” way forward with this. It is a discussion we should all be encouraging, and yet there are those who would rather not have it.
Here is the FRA’s opinion on the subject matter: http://www.apta.com/mc/hsr/previous/2011/presentations/Presentations/Implementing-High-Speed-Rail-in-the-United-States.pdf
– The FRA believes that High Speed Rail equipment must be operationally compatible with existing Tier 1 service (slides 25 & 27)
– HSR systems height and width need to be compatible with North American equipment to allow sharing of tracks (slide 16)
– The Rail Safety Advisory Committee (RSAC) Engineering Task Force (ETF) is developing criteria for HSR train sets (width, floor height & ADA compliance strategies) (slide 22)
– The FRA believes that too many changes in configuration can upset the balance and cause unexpected safety issues (slide 33)
[…] Caltrain Considers Train Cars With Two Set of Doors for Lower Platforms and CAHSR Platforms (GC) […]
Where in the world uses such “SPECIAL” EMU?
Caltrain made right choise when they decide to non-FRA EMUs because of lower operating/maintenance cost and cheaper rolling stock price from the “off the shelf”.
This 2 different door EMU completely offset those cost saving.
Caltrain need to consider
(1) How much is the additional cost of such special rolling stock?
(2) What is the cost of maintain 2 sets of door and conversion?
(3) What is the cost of raising platfrom at the time of electrification? (It can be raise the platform with the speed of curing concrete.)
I am not saying taking HSR money and run with it. It is important to note that even if Caltrain and HSR were to share tracks, they don’t have to share everything else.
This would be no different than roommates sharing the same room with separate beds. If the house has more rooms, you would want to sleep in separate rooms when the opportunity is available, especially if the roommate snores when sleeping. There are cost and benefits trade offs between various levels of sharing. If you share just enough, you can save cost and maintain a reasonable level of comfort and privacy, sharing beyond that would create conflicts.
If HSR has their way, they probably wouldn’t want to share anything with Caltrain, but sharing tracks is reasonable enough. HSR doesn’t really want to share platforms either, so that’s why they would do so if Caltrain runs that their height, and that’s why you advocate Caltrain to run at that height. This is kind of like a shotgun wedding between unwilling parties, but one party (Caltrain) has to sacrifice way more than the other. The hope is that they can share the same bed, but such marriage may not last.
The only place that has a real space issue is Transbay. All the other talk about Millbrae and San Jose is a pure distraction.
With regards to Millbrae and San Jose, it would actually be irresponsible not to design and make room for extra platforms, even if initial service level theoretically allows HSR to share platforms with Caltrain. The advantage for platform compatibility is relatively minimal in Millbrae and San Jose. Also, platforms doesn’t take that much more room compared to supporting facilities like parking. Suburban HSR stations with 6 tracks in China take about the same footprint or less than a 2 track suburban BART station. What makes other mid-peninsula HSR stop unlikely is that the cities don’t have room for facilities like parking.
Caltrain cannot raise platforms at the same time as electrification because it has not received environmental clearance, and the environmental clearance for electrification doesn’t include raising platforms.
Andy,
Is it subject to enviromental clearance for raising platfrom? Since platform is already exist with 8 inch. From 8 to 25 or 50 inch, there will be no impact on “Enviroment”.
As long as the rail car uses standard sized wheels, there will be sections of the car at ~50” OTR. So it make sense to do level-boarding at ~50”. Any platform height higher than today’s 8” requires waivers, so why not at a height that ensures HSR compatibility.
As for speed of boarding, it is going to be faster for people who sits in the upper level since they will climb/go down one less flight of stair.
Caltrain Electrification RFP already specifies loading gauge the same as CHSR, so Caltrain can order rail cars that’s the same width as CHSR, enabling 3+2 sitting arrangement if necessary.
Whether or not there’s enough land for extra platforms is a different question than whether it’s a better use of the land, and a better passenger experience, to have a larger and more spread out station.
A policy statement that all regular service Caltrain trains (except perhaps ballpark specials) should serve Transbay is not a “positional statement” opposed to other interests. Transbay is expected to be the highest ridership station on the Caltrain system. All regular service trains should serve Transbay, not only for the interest of the City of San Francisco and the Transbay Authority, but for all other stakeholders one the line – employers whose employees live in SF and East Bay, cities whose residents work in downtown SF, everyone who wants more car traffic off the roads. It’s the best value out of the billions that will be spent by taxpayers.
This is crazy idea but can we convert Gallery car to 50 inch by eliminating step?
This can start few month before the electrification complete. Caltrain schedule to be simplyfied to local (Gallary Car) and Express (Bombardia) only.
SF and SJ have lots of track for building different platform height. Milbrae can utilize platform 4s and westside of platfrom 5. 22nd and Palo Alto have long enought platfrom to have different height. Choose Hilsdale or Redwood City, Mountain View or Sunnyvale to constract two different height of platform. Some station temporary lost express service but increase local train to offset the convenience.
There are not enough Bombardia car for operating all express train. So, we can borrow from ACE’s bombardia car for this period and let them use our Gallery car for this time period.
Gallery cars have floors at 45″ ATOR.
[…] 78 to 188 seats will be lost per six-car train. Agency officials have clarified that this need not occur until the High Speed Rail stations are actually built, because trains could be built without the upper set of doors, initially, in order to temporarily […]
1) How do you know that there’s no impact on environment until it is studied? and according to the law the environment includes anesthetics, trees, and impacts on historical structures.
2) It doesn’t make sense because it doesn’t work well for Caltrain customers, especially the cyclists and disabled, and the height of such platform is not wanted by Caltrain communities.
I understand the desire for 50″ platform assumes that HSR is a very important component, but that’s not something necessarily shared by Peninsula communities. Many of them don’t want HSR (or at least with these pre-conditions), and some of them would not want electrification and other improvements because of their association with HSR.
For HSR and other improvements to go forward, there should be no more “mandates” besides what’s already approved and agreed in the blended plan. There are workarounds with separate platform heights.
3) If we are to build a new airport, that new airport should’ve enough land to have multiple runways, even if the short term demand only requires one. It would be bad planning practice to argue that common height would result in smaller station foot print. We are not talking about a 10 track shared terminal versus a 20 track separate terminals, we talking about 2 platforms versus 3 or 4 in case of Millbrae. That increase in 1 to 2 platforms would significantly improve flexibility and capacity, and the increase in footprint would be marginal.
As for passenger experience, most high frequency corridors don’t have shared ticketing, because HSR service is considered first class/premium service compared to regional/transit like service. In California, shared ticketing between intercity/regional rail happen in areas with infrequent service and they all have some sort of restrictions.
It is likely that California HSR would have multiple classes service common in Amtrak, European, and Asian HSR. It is likely that HSR will implement reservation based ticketing common with other HSR systems as well as airlines.
4) I don’t disagree with the concept that there should be as many trains serve Transbay as possible, but I don’t see any suggestion made by staff at the three agencies that somehow a common height platform would allow all Caltrain serve Transbay, or that somehow separate platforms would not allow most if not all trains to serve Transbay. There’s no simulation of any sort made to suggest either or.
I can say that in the New York area, not all LIRR and NJT lines serve Midtown Manhattan. Of course everybody know Midtown Manhattan is where all the action is. But both LIRR and NJT maintain terminals just outside Manhattan for extra peak trains and particularly diesel trains.
There’s a difference between having as many trains serving Transbay versus forcing all trains to serve Transbay because alternative terminal is loss to development.
5) Clem Tiller should highly recommend this:
http://rrpicturearchives.net/showPicture.aspx?id=3883719
It seems like there are other solutions to the problem of different platform heights that should be looked at.
In Switzerland, I’ve encountered bilevel coaches that you board from high-level platforms by descending a short ramp inside the train. It seems like you could have a tilting ramp for compatibility with two platform heights. It’s not quite level boarding, but it could be close enough for bikes or wheelchairs. Certainly a lot more convenient than getting a bike or wheelchair up and down steps on a moving train (obviously not possible for a wheelchair–this issue doesn’t seem to have gotten much attention).
Another idea: could the few stations that actually will get service from both CalTrain and HSR (Diridon, Transbay, Millbrae–any others) have adjustable *platforms*? It seems kind of goofy, but really, why not? The platforms would only have to be adjusted a few times a day when shared tracks switched between serving HSR and local trains. And this would just be something custom installed in a handful of locations, which seems simpler than requiring every single train car to be custom.
Anyway, I think the Wheelchair/ADA issue should be a dealkiller for the multi-level-door-car solution.
Metra Electric will be good example for Caltrain. I am proposing 48″ height as the compromaise between HSR and Gallery car.
This proposal is under assumption of Caltrain to purchase 50″ EMU.
Step-1 Change Caltrain existing consist to
3-Gallery Car + Locomotive + 3-Bombardia Car or
5-Gallery + Locomotive. (North to South)
Step-2 Construct three 48 inch mini-high platform north side of platform at all the station.
Step-3 Remove step from northern three Gallery cars of all the consists.
Step-4 Raise 3-cars long northern side platform to 48″.
Step-5 Extend 8-in platform to south for 3-cars long but this is limited at express stop stations (22nd, Milbrae, Hilsdale, Redwood City, Mountain View and Sunnyvale). SF, Palo Alto, SJ and Tamien are already long enough.
Step-6 Electrification complete. Start using 3-car EMU with 50″ floor height for local service only and 3-car consist.
Increase local train frequency to handle capacity. (Operate with single conductor)
Express continue to use diesel until platform conversion. Change the consist with 3-Gallery (48″) + 3-Gallery (8″) + Locomotive
All the Bombardia to use Tamien-Gilroy service only.
Step-7 Build three 48″ mini high for southside of platform at express stops. Extend 48″ platform at all other stations.
Step-8 Remove steps of south side of Gallery car.
Step-9 Raise all the remaining 8″ platform to 48″
Step-10 Retire Gallery/diesel train and full replacement to EMU (3 or 6 car consist)
Re last sentence in Andy Chow’s comment: “What makes other mid-peninsula HSR stop unlikely is that the cities don’t have room for facilities like parking.” Potential solution is to build parking over the tracks — use ultra-light-weight, high-strength structures like Bosch Captive Column geometry to make it cost-effective.