Firms annoyed with STM over iBus contract
The STM has been considering bids for the contract to build its iBus system, a GPS-based network that will improve on its static bus schedules by telling passengers exactly where the bus is and how long they will have to wait. I wrote a couple of short pieces about the idea earlier this year, including the rationale for the rather hefty $200-million price tag for providing a kind of system that already exists and in fact is in regular use already by the Laval transit system.
Now there’s news that the firms bidding for the contract are unhappy with how the process was done and that it’s a German company called INIT that has been chosen even though it wasn’t the lowest bid. It is, however, less than half the initial $200 million mentioned, which makes me wonder what’s been left out of the spec.
Tux 10:01 on 2012/09/06 Permalink
As they proved with OPUS, STM management is ill-suited to projects involving technology. I’ve said it before… get some ETS students on this project. They’d deliver a great system under-budget and we’d be supporting our youth/offering real engineering experience at the same time.
Kate 10:03 on 2012/09/06 Permalink
That’s a really good idea.
Bert 10:09 on 2012/09/06 Permalink
Who cares where the bus is? It arrives on a schedule and that is it, plus or minus a couple of minutes who cares?If you need to schedule your bus travels down to minutes you should just leave earlier or seek some other form of transport.
walkerp 10:27 on 2012/09/06 Permalink
Because it is so often not on schedule. If you knew where they were precisely, you could plan when to leave the house, work, etc. to catch it right on time with no waiting, which makes for more efficiency for everybody.
Kate 10:33 on 2012/09/06 Permalink
The STM knows if they’re going to make bus travel as attractive and convenient as possible for more users, they have to serve a lot of impatient people. Anyway, lots of transit systems have benefited from adding this feature. Ours will too.
Bert 10:45 on 2012/09/06 Permalink
If people are impatient, and if the buses are late, maybe the money would be better spend making the buses run on time. I know it sounds like a crazy idea, but fixing the PROBLEM should be the priority, no?
jeather 10:47 on 2012/09/06 Permalink
A bunch of engineering students are not the ideal group to build this kind of system. I’ve used projects that are done that way and — well, no. They don’t have the experience to do it properly, and we’d get a system that is sure to infuriate everyone.
cheese 11:03 on 2012/09/06 Permalink
A system of this kind is a very good idea, London (England) has displays in the bus shelters which show a rotating list of up to ten buses which will arrive at that location next. It includes the bus number, expected wait time, and final destination for each bus.
Very good system which can also be queried by mobile phone (via sms and more sophisticated smartphone means), it allows one to not just plan your usual bus journey but decide if it is better to walk or take a different bus based on the current timing.
There is no way to make the busses “run on time” all the time unless they had dedicated lanes for their entire routes.
Hoedic 11:07 on 2012/09/06 Permalink
@Bert: Information system is part of solution. Even if STM does its best to improve the situation, you will never go over congestion, even with dedicated lines. So a realtime tracking system help people to plan and avoid them waiting outsite when it’s -20.
Added to that, it’s more than just data for the customer: it’s also data for the STM and for analysis, it’s a fleet management tool. That can be very useful at several levels.
Blork 11:27 on 2012/09/06 Permalink
I would LOVE to know where the buses are! There are times when it doesn’t matter (such as when you’re already waiting and there are no alternatives), but many times you have a choice of buses and if you know one is close and the other delayed, that will help you choose which to go for. Or sometimes it’s a choice between the bus and the Metro. Again, if you know the bus is only two minutes away, versus 12 or 15, that can affect your choice.
Ian 13:12 on 2012/09/06 Permalink
@Bert – if you’re waiting outside in the winter, a few minutes makes a big difference. FWIW most bus delays are caused by traffic issues which are irregular & unpredictable – and throwing more buses on the line won’t fix that. If you know the bus is going to be 15-20 minutes late becasue of a traffic jam or snowy roads or whatever, in many cases you can walk to a more frequent line, leave work a bit later (or earlier), walk to the nearest metro instead or just go grab a coffee to pass the time. As a regular bus user, I think this is a GREAT idea.
SN86 14:23 on 2012/09/06 Permalink
Copy exactly what the STL (Laval) has already done with their GPS system.
Raymond 15:23 on 2012/09/06 Permalink
OK juste pour épater la galerie avec mes talents d’ingénieur d’estrade: crowdsource this fonction aux usagers eux-mêmes, avec l’ubiquité des cellulairess sur les bus…
There’s even an app for that… http://www.waze.com
Processing of aggregated locations could distinguish car form bus users (and moving patterns too… d’autant plus que le tracé des bus est déjà connu).
That’s how Google map does it. And if you don’t want to use GPS data, you could always triangulate cellular signals.
meezly 15:53 on 2012/09/06 Permalink
Since I’m in my late pregnancy and have no car, I really depend on the bus every day to get to work and to my doctor appts. I just can’t hop on a Bixi or walk 20 min in hot, humid weather.
What kills me is catching a bus when in transit. I don’t know how many times I got off the metro to wait for my bus (along with several people)… and it never showed up at its scheduled time. Then I later observed that sometimes some buses actually arrive & depart several minutes BEFORE their scheduled arrival time. I have written to the STM requesting that buses arriving before their scheduled time should really wait at the stop, but of course nothing has come of it.
This isn’t just one particular bus, but 3 buses that have shorter routes that only run 2-3 times an hour during rush hour (the 715, 61, and 74). So if I miss one I have to wait another 15-20 min, or hustle to another stop to catch another bus.
I would rather a bus be a few minutes late than leave too early. So knowing if a bus had actually missed its scheduled stop would help a lot.
ant6n 20:39 on 2012/09/06 Permalink
@meezly
Agreed. (Very much so)
STM getting a good contract is good news, actually. I was under the impression they were going to get some in-house solution; using something more or less off-the-shelf is the best approach.
Tux 13:31 on 2012/09/07 Permalink
@Jeather – Disagree! This is not a complicated project. You need basically 4 main components. A GPS receiver on the bus (dirt cheap, easily sourced) some kind of transmitter to send the data to home base (make a deal with Videotron or Bell on the data rates and use GPRS transceivers – also cheap and easily sourced) a microcontroller to manage communication between the GPS and the radio, and a central software system to receive and interpret the data and format it for public consumption (a where’s my bus? website/app) I’m a programmer by day and an electronics hobbyist by night, and rolling something like this would not be outside of my capabilities. We do not need some overly expensive pre-packaged “turn-key solution” provided by some big European company. That was the mistake we made with Opus.
Tux 13:46 on 2012/09/07 Permalink
@Jeather Also the implication that students can’t create a solution that’s better and cheaper than a big company because they’re students is just condescending. People are people, some are older, some are younger, some are smarter, some are dumber. Whether or not they are employed by a company is largely irrelevant. Big companies release crappy products all the time. Have you SEEN the kind of stuff ETS students build? Submarines and solar powered UAVs and all manner of cool stuff. You give them a spec document where the operational requirements are laid out, at the end you verify that the tech meets the spec… boom! New system, under-budget, rapid development cycle, and new features rolled out quickly and efficiently! You’re not waiting on some German company with a bunch of other interests to find time to release a Montreal-only patch to their internationally sold product, you pay a local programmer (student or not!) to do it and get it done.
ant6n 14:53 on 2012/09/07 Permalink
You gonna have to build like 2000 of those, and install them, and what’s the reliability going to be like?
Heb3rt 15:19 on 2012/09/07 Permalink
@Kate Here’s the answer to your question about he 200M$ initially announced taken from the media release of STM:
The three-and-a-half year contract was awarded as part of the i-BUS project, which also includes other investments related to expenses for the project bureau and various analyses, as well as those pertaining to the establishment of an operational centre and the integration of infrastructure and interfaces with the STM’s other systems. Estimated at $200.6 M, this major project was approved by both the City of Montréal and Montréal Urban Agglomeration in October 2010.
Tux 15:34 on 2012/09/07 Permalink
@ant6n It’ll be as reliable as the GPS reception, cell phone network and bus power supply (battery). You build a few reference boards – testers – by hand. Then once you have a design that tests reliably you have the boards fabricated and mounted in weatherproof enclosures on top of the bus. Like I said, vehicle tracking systems aren’t rocket science. Also there are definite societal and economic benefits to keeping the job in-province and to getting students to participate. Taxpayers get a cheaper system that’s just as good (and even if it is not initially as good as a more mature product, upgrades could be much speedier with only a few dedicated local people working on the hardware and software)
ant6n 17:06 on 2012/09/07 Permalink
I don’t understand why students should be used as a cheap (below market-value), unreliable labor force, rather than buying off-the-shelf tech. If you have many bids, then you should get market price, i.e. a price that corresponds to the complexity of the problem. If you pay much less by employing people below their qualifications, you may get what you pay for.
Kate 10:12 on 2012/09/08 Permalink
Heb3rt, thanks for the clarifications.
My main concern about this project is that, like others, I can see Montreal reinventing the wheel. We like to do that, we reinvent the Velib’, we pay thousands for nifty-looking custom trash bins rather than buying stock. Some of the rationale has to be job creation, but not all of it. If a stable, reliable GPS bus system exists (and Laval is using it) it’s a mystery to me why Montreal needs to start again from scratch.
Raymond Lutz 13:45 on 2012/09/09 Permalink
@ant6n:
Are you living in an econ-101 textbook? 8-)
Avant de citer Adam Smith, lisez:
http://en.wikipedia.org/wiki/Invisible_hand#Abusing_Smith.27s_statement_of_an_invisible_hand
Tux 15:11 on 2012/09/10 Permalink
I don’t understand why students should be used as a cheap (below market-value), unreliable labor force, rather than buying off-the-shelf tech.
Okay, again, just because they’re students and not employed by a company, does not mean they are unreliable. Where do you get this stuff? As for cheaper, I am thinking the total cost of rolling our own system would be cheaper than any massively marked up commercial device (even if you paid the students the market rate for their specialties), and having access to circuit diagrams and source code future-proof the project, whereas when you buy commercial tech you are more or less at the mercy of the parent company when you need parts, upgrades, etc. (And they charge you through the nose because they can) This same thing happened when the company that built the original fare system hardware for the STM went out of business and they were forced to hoard parts and do their own maintenance without support.
If you have many bids, then you should get market price, i.e. a price that corresponds to the complexity of the problem. If you pay much less by employing people below their qualifications, you may get what you pay for.
We paid a market price for Opus, a ‘mature’ system built by ‘experts’ and look what it has gotten us. Most of the promised features are still undelivered, and the system has no real advantage over the old system in terms of security or speed.
A diploma’s a piece of paper buddy – I guarantee you that there are gainfully employed, diploma’d electrical engineers who couldn’t hold a candle to some of the talent being turned out by our schools. You put too much faith in officialdom. Plus, you put a local student on this, they’re going to eat it up, they’re going to be passionate about the project. They’d be building a system that would be used in their hometown, and they’d get to brag about it! If I’d been offered an opportunity like that as a youngster I would have jumped at it, and I would never have settled for turning out a flawed product.
Also, again, this keeps the money in-province (arguably more important than getting the cheapest price), and is of direct benefit to up-and-coming engineers.