"Leap" into the unknown: The feedback thread

cdebru

MYOB wrote: »

Ticket machine software can now easily be implemented on mass-market hardware, as can nearly everything else.

So an android phone would make a good bus ticket machine ??

L1011

cdebru wrote: »

So an android phone would make a good bus ticket machine ??

As posted, they're using Android tablets for other Leap operators already.

However, you appear to have failed to notice that the comparison was purely one to show that a defence of the utterly pathetic specs, even when introduced, of that model of Wayfarer on cost grounds is not valid.

cdebru

MYOB wrote: »

As posted, they're using Android tablets for other Leap operators already.

However, you appear to have failed to notice that the comparison was purely one to show that a defence of the utterly pathetic specs, even when introduced, of that model of Wayfarer on cost grounds is not valid.

No I accept that the current ticket machine is wholly inadequate and that should have been obvious when these machines were purchased, I just don't see the point in comparing it to a 40 euro android phone.

Like I said before the people responsible for acquiring this hardware and implementing software updates do not use the service and there is no channels for feedback from those that do use them, hence I don't believe they are aware they are not suitable, and that is why they are being rolled out on bus eireann buses.

AlekSmart

antoinolachtnai wrote: »

The NTA supplied equipment for private operators is basically a Motorola Xoom tablet.

Dublin Bus, not the NTA, chooses electronic ticketing machines for Dublin Buses. This might change but this is how it is and how DB wanted and wants it.

The stuff about crypto is not really correct. The crypto is done on the NTA supplied card reader, not on the ticket machine. There is however the marvellous world of RS485 serial bus communications to contend with. It is tricky to integrate but this is really an issue for DB and its supplier.

I suspect that BAC's specification of the TGX150 was down to the machines eminent suitability for BAC's own operations,as can be seen by it's excellent performance up until Leapcard functionality was bolted onto it.

With the ITS one of the NTA's Flagship elements stretching back over a decade,it still escapes me as to WHY the NTA did'nt nail down all of this specification/compatibility from Day 1.

This was never about some huckster peddling new machines or programmes,it was about a NATIONAL Integrated Ticketing Programme,with the main operators all being part of the same Semi-State Group.

This gave the ITS Team an incredible advantage in terms of getting a single operational platform up and running well in advance of Leapcard Day.

Why this advantage remained unexplored remains shrouded in mystery,when a single Ministerial Order could have DIRECTED the CIE companies to get their house into Leapcard shape PDQ.

If,as Antoin sez,its down to BAC not being fully compliant,then full marks to BAC management for being able to comprehensively frustrate Government Transport Policy in such a fashion whilst remaining in their jobs....:o

The current situation is beyond belief and rational explanation,but is shaping up to be a textbook example of how NOT to implement ANYTHING !!!

I noticed today that when I boarded a 41, the passenger in front of me used their Leap and got the long delay, whereas when I put mine on it, it registered right away. Why would this be the case?

AlekSmart

devnull wrote: »

.

It has only 1mb of memory, which is it's first limitation, and it is based around a chip which was introduced by Intel in the mid 1980s (The 386), which has been discontinued for nearly seven years now. Whilst it can do simple read write and ticket printing duties on multi day, throw away tickets, the extra functionality that leap requires places it's slow, outdated processor under strain.

Add in the fact that the Leap Cards are encrypted to a high level, which is always going to be required on such a system where someone keeps a card for a long time and reuses it, to prevent fraud and card cloning that can and will happen if the security is not strong enough, and there's just too much for that hardware to do.

There is a much faster version of this machine that was introduced almost six years ago to overcome some of the problems as listed above with the existing machine. It has over 100 times more memory, much more modern technology and a much faster processor.

Really the NTA need to try and get their finger out and give some money up to try and replace these older machines or at least contribute something to them. The only other workaround is to dumb leap down, or to reduce the encryption on the cards, the former will reduce the features that LEAP can offer, and the second will make card cloning and fraud much easier. So neither are desirable.

Whilst Antoin,elsewhere casts some doubt on the relevance of the encryption issue,I still have serious doubts concerning how the entire ITS programme has been stewarded from Day 1.

The question I pose to Devnull,who has some interest/ability in Software/Systems,is that if,six years ago you were presented with addressing the TIM/Leapcard integration issue,would you have progressed as the NTA have done ?

If not,can I ask what exactly you would have done differently...?

AlekSmart

Karsini wrote: »

I noticed today that when I boarded a 41, the passenger in front of me used their Leap and got the long delay, whereas when I put mine on it, it registered right away. Why would this be the case?

This is the now standard Leapcard situation.

It appears that the initial Leapcard presentation enables the Leapcard specific capability,with succeeding Leapcard presentations being carried out within that original envelope.

However,should the Leapcard series be interrupted by a Cash Transaction or remote BAC Card validation,then the TIM will leave the Leapcard mode and have to re-establish it from scratch the next time a Leapcard is presented.

Victor

Karsini wrote: »

I noticed today that when I boarded a 41, the passenger in front of me used their Leap and got the long delay, whereas when I put mine on it, it registered right away. Why would this be the case?

The suggestion is that it 'goes to sleep'. I suspect it is reloading the software.

cdebru

Karsini wrote: »

I noticed today that when I boarded a 41, the passenger in front of me used their Leap and got the long delay, whereas when I put mine on it, it registered right away. Why would this be the case?

Perhaps you might make DB aware that you have witnessed this event, maybe they will believe you.

brownej

Karsini wrote: »

I noticed today that when I boarded a 41, the passenger in front of me used their Leap and got the long delay, whereas when I put mine on it, it registered right away. Why would this be the case?

cdebru wrote: »

Perhaps you might make DB aware that you have witnessed this event, maybe they will believe you.

I don't know how the powers that be and the drivers have not noticed this. I do not live in Dublin but I use a leapcard when I am there and the last two times I used my card on the bus this situation occured. The person in front of me takes ~15s to perform a leap transaction and then mine completes immediately.
The first time I noticed this was on the 10th of August so the problem has existed for a few weeks.

markpb

brownej wrote: »

I don't know how the powers that be and the drivers have not noticed this.

The drivers do know it and I'd be surprised if the management didn't know either but there's not a lot they can do except hope that Wayfarer write better software or that they find a pot of gold at the end of a rainbow and use it to buy new ticket machines.

pclive

Its all well and good discussing this issue here but has anyone taken the time to write to leapcard/dublin bus or the minister to complain about this.

I received a reply back from leapcard stating that they are aware of the problem and that a software upgrade is being prepared to speed up the transactions. This softare upgrade will be rolled out across the Dublin Bus fleet in the coming weeks.

Why not take the time to email the minister at minister@dttas.ie highlighting the issues with the leap card roll out.

devnull

brownej wrote: »

I don't know how the powers that be and the drivers have not noticed this. I do not live in Dublin but I use a leapcard when I am there and the last two times I used my card on the bus this situation occured. The person in front of me takes ~15s to perform a leap transaction and then mine completes immediately.
The first time I noticed this was on the 10th of August so the problem has existed for a few weeks.

That's because for the next person the software is already loaded in memory by the person who was in front of you and therefore it doesn't need to load it again.

The problem with the lack of memory on the machines is it cannot store much in there and is slow at loading into it, so when it goes from LEAP mode to another mode, it has to then reload that software and clear LEAP from memory, then for LEAP it needs to reload again.

If they had more memory in the ticket machines, it would be able to store both applications in it at once, and a faster processor would also help since it would speed up the loading of such software to/from the memory.

L1011

cdebru wrote: »

No I accept that the current ticket machine is wholly inadequate and that should have been obvious when these machines were purchased, I just don't see the point in comparing it to a 40 euro android phone.

The alternative unit as used by other operators is a mass-market Android device. Did you fail to notice that bit?

cdebru

MYOB wrote: »

The alternative unit as used by other operators is a mass-market Android device. Did you fail to notice that bit?

Available for €40 ? You also need a card reader, and a printer to print of tickets not to mention the software to deal with the DB staged fares model, the other operator is Wexford bus, hardly comparable to the high usage frequency of DB is it ? how long do you think an android tablet would last on a Dublin Bus ?

Again the price of a subsidised or unsubsidised mobile phone is irrelevant.

bk

cdebru wrote: »

Available for €40 ? You also need a card reader, and a printer to print of tickets not to mention the software to deal with the DB staged fares model, the other operator is Wexford bus, hardly comparable to the high usage frequency of DB is it ? how long do you think an android tablet would last on a Dublin Bus ?

Again the price of a subsidised or unsubsidised mobile phone is irrelevant.

The tablet used now would cost about €350 (Nexus 10) for a top quality model. But you could get cheaper chinese models for €150

I believe the company who developed this system charges €1000 per bus for the entire system, including tablet, mounting, printer, card reader * and all the back end systems including ability to track bus locations, send messages to all or individual drivers, update timetables, etc.

€1000 per bus is very reasonable for all of this. The DB ticket machines easily cost more then this due to their property/custom nature.

I'd have little concerns about robustness. In my experience touch screens are far more robust then physical buttons. With the exception of dropping and breaking a screen (which obviously wouldn't be problem on a bus with a mount) I've never seen a touchscreen break on a phone or laptop. I've seen the physical buttons like the iPhone home key on phones/tablets break or the volume buttons, but never a screen.

Also the cost of replacing a broken unit would be very reasonable. Last year when we first started talking about this, a top of the range tablet cost €500, today it is €350, next year it will be €200. A wayfarer machines costs far more then that to replace due to it's custom nature.

Actually Wexford Bus have started using this new ticket machine after it was successfully trialled with Matthews Coaches, so it looks like the trail was a success and I imagine it will only be a matter of time before it is also trialled and then rolled out with DB.

Interest to note, that the NTA are actually paying for these ticket machines for the private operators.

cdebru

bk wrote: »

The tablet used now would cost about €350 (Nexus 10) for a top quality model. But you could get cheaper chinese models for €150

I believe the company who developed this system charges €1000 per bus for the entire system, including tablet, mounting, printer, card reader * and all the back end systems including ability to track bus locations, send messages to all or individual drivers, update timetables, etc.

€1000 per bus is very reasonable for all of this. The DB ticket machines easily cost more then this due to their property/custom nature.

I'd have little concerns about robustness. In my experience touch screens are far more robust then physical buttons. With the exception of dropping and breaking a screen (which obviously wouldn't be problem on a bus with a mount) I've never seen a touchscreen break on a phone or laptop. I've seen the physical buttons like the iPhone home key on phones/tablets break or the volume buttons, but never a screen.

Also the cost of replacing a broken unit would be very reasonable. Last year when we first started talking about this, a top of the range tablet cost €500, today it is €350, next year it will be €200. A wayfarer machines costs far more then that to replace due to it's custom nature.

Actually Wexford Bus have started using this new ticket machine after it was successfully trialled with Matthews Coaches, so it looks like the trail was a success and I imagine it will only be a matter of time before it is also trialled and then rolled out with DB.

Interest to note, that the NTA are actually paying for these ticket machines for the private operators.

So €1000 which is a long way from a €40 android phone that's without any links supporting your pricing claims. Also in general most of these companies make the money in ongoing support they won't allow you to just pop in a replacement tablet.
The CCTV system is an example it is was just a simple 2.5" hard drive, but when they went south which hard drives do, they had to be replaced by the supplier at multiples of the cost of a larger hard drive. Otherwise they won't support the system at all.

Also what is the cost of the newer model wayfarer ? If we are doing a comparison.

Personally I don't think tablets are durable enough for city operations.

AlekSmart

brownej wrote: »

I don't know how the powers that be and the drivers have not noticed this. I do not live in Dublin but I use a leapcard when I am there and the last two times I used my card on the bus this situation occured. The person in front of me takes ~15s to perform a leap transaction and then mine completes immediately.
The first time I noticed this was on the 10th of August so the problem has existed for a few weeks.

Brownej,this situation was "Noticed" from T+1 second on the morning after the NTA "Software Upgrade".

This "Upgrade" was,it appears,to prepare for the then imminent introduction of Daily/Weekly Capping on Bus Leap Transactions.

Whatever software alterations were made were obviously NOT tested in a live environment before being signed-off for full issue.

I would contend that this points towards a significant problem within the software development and application area of Leapcard,and equally threatens the very principle of "Integration" which Leapcard is meant to be the Flagship of.

THE WAYFARER WAS PERFORMING NORMALLY BEFORE THE SOFTWARE UPGRADE.

THE WAYFARER IS NOT PERFORMING NORMALLY AFTER THE SOFTWARE UPGRADE.

QUESTION: Is there a connection between the Newly Developed Problem and The Software Upgrade ? :rolleyes:

bk

cdebru wrote: »

So €1000 which is a long way from a €40 android phone that's without any links supporting your pricing claims.

You are ignoring the original point I was making about the €40 smart phone.

I was showing how cheap a powerful processor, ram and rom have become now for off the shelf devices.

Of course a more robust large tablet along with all the supporting infrastructure needed for a bus is going to cost more. But certainly inline if not significantly cheaper then the wayfarer.

You'll just have to trust me on the €1000 figure, it was originally in the video on the Avego website, but that video is gone now unfortunately. The video also showed the very impressive backend systems.

cdebru wrote: »

Also in general most of these companies make the money in ongoing support they won't allow you to just pop in a replacement tablet.
The CCTV system is an example it is was just a simple 2.5" hard drive, but when they went south which hard drives do, they had to be replaced by the supplier at multiples of the cost of a larger hard drive. Otherwise they won't support the system at all.

Yes and that is the beauty of a Android tablet system over the Wayfarer.

The Wayfarer ticket machine can only be bought from Wayfarer and the software on it only developed by them. They basically have a complete monopoly and you can bet they squeeze every cent out of DB for any change made.

By using standard off the shelf android tablets instead, it makes it far easier to change supplier. Any android table can be used, the software written for todays android tablet can just be moved over to a new one tomorrow.

Also as the whole system is open source, much easier to bring in a new company to update and change the software in future.

Another point is that as a software engineer myself, I can assure you that it is vastly quicker and easier to develop software in Java on a high end android tablet with lots of memory and CPU then it is to develop in a low end embedded systems language on a device with very restricted CPU and memory.

Developer costs per hour would be by far the biggest costs in any system like this, far outstripping even the hardware costs. They are massive advantages to
using an android system like this to keep development costs down.

devnull

AlekSmart wrote: »

THE WAYFARER WAS PERFORMING NORMALLY BEFORE THE SOFTWARE UPGRADE.

THE WAYFARER IS NOT PERFORMING NORMALLY AFTER THE SOFTWARE UPGRADE.

QUESTION: Is there a connection between the Newly Developed Problem and The Software Upgrade ? :rolleyes:

Yes, the kind of software that the Leap card requires to run is beyond the capability of the hardware that it is running on, that clearly is the case however much it is dressed up by any party.

The Leap card is a far more sophisticated product than any of the types of tickets DB offer so would be more intensive for any ticket machine, for several reasons and that is the way it should be.

The fact is, would you go and try and get a 15-20 year old computer to run softtware that has come out in the last few years, then blame the software as to why it doesn't work properly?

brownej

bk wrote: »

You are ignoring the original point I was making about the €40 smart phone.

I was showing how cheap a powerful processor, ram and rom have become now for off the shelf devices.

Of course a more robust large tablet along with all the supporting infrastructure needed for a bus is going to cost more. But certainly inline if not significantly cheaper then the wayfarer.

You'll just have to trust me on the €1000 figure, it was originally in the video on the Avego website, but that video is gone now unfortunately. The video also showed the very impressive backend systems.



Yes and that is the beauty of a Android tablet system over the Wayfarer.

The Wayfarer ticket machine can only be bought from Wayfarer and the software on it only developed by them. They basically have a complete monopoly and you can bet they squeeze every cent out of DB for any change made.

By using standard off the shelf android tablets instead, it makes it far easier to change supplier. Any android table can be used, the software written for todays android tablet can just be moved over to a new one tomorrow.

Also as the whole system is open source, much easier to bring in a new company to update and change the software in future.

Another point is that as a software engineer myself, I can assure you that it is vastly quicker and easier to develop software in Java on a high end android tablet with lots of memory and CPU then it is to develop in a low end embedded systems language on a device with very restricted CPU and memory.

Developer costs per hour would be by far the biggest costs in any system like this, far outstripping even the hardware costs. They are massive advantages to
using an android system like this to keep development costs down.

Unfortunately using consumer grade electronics in an always on commercial setting such as dublin bus is leaving yourself open for all sorts of issues.
If we leave aside the physical robustness of the device for constant use on a bus we still have the issue that the manufacturers of these pile em high and sell em cheap (yes even the expensive ones are cheap,as in they have low margin based on selling large numbers) are built assuming a standard consumer usage profile. That usage profile does not include continuous use for 18 hours a day, 7 days a week.
For use on Dublin bus these tablets would need to be modified. The processor speed would need to be dialed back and the applied voltage reduced. This will have a very real affect on the sources of intrinsic failure on the device and should get at least another year of useful life out of the processor. The software (operating system) would need to be stripped down and locked down (no installing angry birds) to minimise on any potential sources of hanging or crashes.
Even doing this you would still be replacing the tablets on a regular basis.

I have also highlighted in your post something which bugs the crap out of me about software engineering attitudes. Throwing faster hardware to get around lazy bloated programming methodologies should not be a solution.

There's nothing wrong with embedded programming. In fact your smart phone/ tablet has numerous embedded processors besides the main CPU.

bk

brownej wrote: »

Unfortunately using consumer grade electronics in an always on commercial setting such as dublin bus is leaving yourself open for all sorts of issues.
If we leave aside the physical robustness of the device for constant use on a bus we still have the issue that the manufacturers of these pile em high and sell em cheap (yes even the expensive ones are cheap,as in they have low margin based on selling large numbers) are built assuming a standard consumer usage profile. That usage profile does not include continuous use for 18 hours a day, 7 days a week.
For use on Dublin bus these tablets would need to be modified. The processor speed would need to be dialed back and the applied voltage reduced. This will have a very real affect on the sources of intrinsic failure on the device and should get at least another year of useful life out of the processor. The software (operating system) would need to be stripped down and locked down (no installing angry birds) to minimise on any potential sources of hanging or crashes.
Even doing this you would still be replacing the tablets on a regular basis.

And yet unmodified iPads are now used all day every day by pilots on airlines!

Also android tablets are being used constantly all day on airlines as entertainment centers.

There is no issue with locking down an android tablet, off the shelf software already exists to do that, it isn't difficult. Likewise there definitely wouldn't be any need to underclock an android laptop. These devices are designed to be very power efficient and to use very low voltages when high performance isn't needed.

Ticketing software would be very simple compared to a game, very low performance, so would always run in a low power state anyway.

AlekSmart

brownej wrote: »

Unfortunately using consumer grade electronics in an always on commercial setting such as dublin bus is leaving yourself open for all sorts of issues.
If we leave aside the physical robustness of the device for constant use on a bus we still have the issue that the manufacturers of these pile em high and sell em cheap (yes even the expensive ones are cheap,as in they have low margin based on selling large numbers) are built assuming a standard consumer usage profile. That usage profile does not include continuous use for 18 hours a day, 7 days a week.
For use on Dublin bus these tablets would need to be modified. The processor speed would need to be dialed back and the applied voltage reduced. This will have a very real affect on the sources of intrinsic failure on the device and should get at least another year of useful life out of the processor. The software (operating system) would need to be stripped down and locked down (no installing angry birds) to minimise on any potential sources of hanging or crashes.
Even doing this you would still be replacing the tablets on a regular basis.

I have also highlighted in your post something which bugs the crap out of me about software engineering attitudes. Throwing faster hardware to get around lazy bloated programming methodologies should not be a solution.

There's nothing wrong with embedded programming. In fact your smart phone/ tablet has numerous embedded processors besides the main CPU.

I would support brownej's view here.

Perhaps it takes some affinity and experience of the actual world of Bus Engineering,but it is a VERY challenging environment for any electrical/electronic device to operate in.

Some folks with long memories may remember the original "Music on Top" experiment of the late 1970's which saw Tape Machines installed under the front long seats of Leyland Atlantean's in Phibsboro Garage and played an endless loop of Larry Gogan and his back catalogue (!!)....within a very short time the hardware simply collapsed,and it was of the highest grade available at the time.

Also of note in on-bus electronics is the issue of a stable voltage supply as the best available DC systems can still suffer greatly from,often fatal,Voltage Fluctuations.

There is a valid reason why Wayfafer and their ilk remain the preferred option for most large operators and it boils down to unit reliability and long life.

The core issue here is what level of software/hardware development did we get for 12 Years and €45 Million+ and why have we arrived at a situation which surely any proficient,competent Software/Hardware IT person could easily have forseen.????

THAT is an issue which,if unaddressed,could well plague the ITS system long into the future !

brownej

bk wrote: »

And yet unmodified iPads are now used all day every day by pilots on airlines!

I believe they are being used as replacments for paper manuals. Not exactly heavy usage.

bk wrote: »

Also android tablets are being used constantly all day on airlines as entertainment centers.

Casual use entertianment would still fit into the consumer usage model. Maybe a little heavier usage but they would have a replacment cycle built into the business model here.

bk wrote: »

There is no issue with locking down an android tablet, off the shelf software already exists to do that, it isn't difficult. Likewise there definitely wouldn't be any need to underclock an android laptop. These devices are designed to be very power efficient and to use very low voltages when high performance isn't needed.

Ticketing software would be very simple compared to a game, very low performance, so would always run in a low power state anyway.

For the mobile grade processors there can be a great advantage to trading reliability off against performance power etc. If you have access to IEEE Explore then I would recommend this recent paper from ST Microelectronics on mobile processors.
As in a Dublin bus application the device would be in an always on state as there is only a couple of minutes at most between stops/uses.

cdebru

bk wrote: »

You are ignoring the original point I was making about the €40 smart phone.

I was showing how cheap a powerful processor, ram and rom have become now for off the shelf devices.

Of course a more robust large tablet along with all the supporting infrastructure needed for a bus is going to cost more. But certainly inline if not significantly cheaper then the wayfarer.

You'll just have to trust me on the €1000 figure, it was originally in the video on the Avego website, but that video is gone now unfortunately. The video also showed the very impressive backend systems.



Yes and that is the beauty of a Android tablet system over the Wayfarer.

The Wayfarer ticket machine can only be bought from Wayfarer and the software on it only developed by them. They basically have a complete monopoly and you can bet they squeeze every cent out of DB for any change made.

By using standard off the shelf android tablets instead, it makes it far easier to change supplier. Any android table can be used, the software written for todays android tablet can just be moved over to a new one tomorrow.

Also as the whole system is open source, much easier to bring in a new company to update and change the software in future.

Another point is that as a software engineer myself, I can assure you that it is vastly quicker and easier to develop software in Java on a high end android tablet with lots of memory and CPU then it is to develop in a low end embedded systems language on a device with very restricted CPU and memory.

Developer costs per hour would be by far the biggest costs in any system like this, far outstripping even the hardware costs. They are massive advantages to
using an android system like this to keep development costs down.

Then you should have mentioned the 1000 euro tablet option really, my point was and is that the 40 euro android phone is irrelevant, if you could replace the ticket machines with a 40 euro piece of kit it would be a no brainer, but even accepting your 1000 euro it is a 1 million euro cost just for the on bus hardware.

Also you neglected the bit about durability, at best I doubt you would get 18 to 24 months out of a tablet in city services conditions. (that's if they weren't stolen before that)

bk

brownej wrote: »

I believe they are being used as replacments for paper manuals. Not exactly heavy usage.

All their paper manuals, checklists, weather reports and airport and route maps, plus they are also used for company email and schedule planning.

Basically they are VERY intensively used throughout the whole of the flight, they contain every checklist that a pilot uses. Pilots use these checklists during almost every phase of a flight.

These are used far more intensively then a bus ticketing system. And obviously are far more mission critical. Yet after intensive testing by the FAA in the US, they are now being widely and very successfully used.

If these can pass FAA endurance testing, I believe they can be used anywhere.

brownej wrote: »

As in a Dublin bus application the device would be in an always on state as there is only a couple of minutes at most between stops/uses.

I agree that the app would be in an always on state, however the nature of the app, would mean that it wouldn't be very CPU or GPU intensive. Anyway in Android, due to it's multitasking nature, many apps run all the time, e.g. the phone and sms apps. A bus ticketing app would be no more intensive then these.

While I agree using a CPU/GPU less intensively would increase it's battery life, the point I'm making is that you wouldn't need to down clock the CPU/GPU upfront as the CPU/GPU will do that themselves anyway as a ticketing app actually would use little CPU/GPU resources.

If you really want to down clock the CPU/GPU for whatever reason it is pretty trivial in android, I can even do it on my rooted android mobile, no big deal.

Alek, power shouldn't be a major problem as such a tablet would have it's own battery, which would act as a sort of UPS, evening out any bus electrical system fluctuations.

As for your point about the popularity of the Wayfarer machines, I believe it is down to the fact that such Android tablets didn't exist 10 years ago, they really are only coming into there own over the last two years and there is perhaps a bit of slowness in moving over to them due to drivers being familiar with them and needing to retrain on a new tablet based system.

Which of course is a fair point, but I believe the advantages of a tablet based system far outweigh this disadvantage.

I've seen many industries that previously used embedded systems rush to off the shelf android and iOS solutions and most have been very successful.

THe first and most obvious is the electronic flight backs now used by pilots that we discussed earlier. Previously many airlines, used proprietary embedded systems eflight bags. They were very large, slow and expensive. Nearly all airlines are moving to the iPad systems now.

Most airlines also previously used embedded systems entertainment systems, most of these are now moving to standard iPads nad Android tablets streaming content from centralised linux servers.

Closer to home I've seen many parking ticket systems for traffic wardens very successfully and happily replace proprietary embedded systems ticket machines with Android based models.

It is happening throughout many industries.

brownej I'd also like to come back to an earlier point you made which I didn't have time to reply on:

I have also highlighted in your post something which bugs the crap out of me about software engineering attitudes. Throwing faster hardware to get around lazy bloated programming methodologies should not be a solution.

There's nothing wrong with embedded programming. In fact your smart phone/ tablet has numerous embedded processors besides the main CPU.

As an old school software engineer with a bit of a taste for optimisation and squeezing every drop of performance out of my code, I fell for what you are saying above.

But looking at it wearing my project manager hat, I have to say from my own experience of the real world, with the exception of the financial services industry where milliseconds can mean millions of euros, I've found that in the vast majority of cases, for most industries, throwing faster hardware at a problem is FAR cheaper then paying developers to optimise their code.

In my experience I would say the same code can be written 6 to 7 times faster in Java or another object oriented programming language on high end hardware (well arm CPU's used in tablets aren't really high end, but compared to embedded systems they are) compared to the same functionality written for an embedded system.

Also the resulting code tends to be less buggy, more reliable and much easier to debug and maintain.

When you are paying very high hourly rates for a developer, that makes a big difference.

bk

cdebru wrote: »

Then you should have mentioned the 1000 euro tablet option really, my point was and is that the 40 euro android phone is irrelevant, if you could replace the ticket machines with a 40 euro piece of kit it would be a no brainer, but even accepting your 1000 euro it is a 1 million euro cost just for the on bus hardware.

But my earlier point wasn't about the cost of the ticketing system, is was to show how ridiculously outdated and underpowered the ticketing machines used by DB are.

1 million euro in the greater scheme of things is very reasonable. Compared to the 50 million that has already been spent on Leapcard, it would be money well spent IMO.

It would fix the awful delays with leap card validation that plague the DB ticket machines and which I believe if not urgently fixed, will lead to the failure of the leap project.

It would also allow for the enabling of a great deal of functionality that the current ticketing machines can't handle.

For instance it would allow for online top-ups to be applied on buses.

It could also power many other onbus systems, such as bus location screens and CCTV. Allowing DB to save money by replacing multiple systems with jsut one system.

cdebru wrote: »

Also you neglected the bit about durability, at best I doubt you would get 18 to 24 months out of a tablet in city services conditions. (that's if they weren't stolen before that)

No, I honestly believe tablets are more durable then machines that use physical buttons. There is simply less to break. With the exception of dropping
(not a problem on a mounted bus system) I've never once seen a touch screen break on any tablet or phone. I've seen the physical buttons break (e.g. iPhone home buttons) but never a touch screen.

Again the FAA wouldn't have approved iPads for use in commercial aircraft if they weren't durable. The FAA have very strict and intensive testing of durability of all electronics used on an aircraft.

The stealing comment is just silly, obviously they aren't just going to bluetack the tablets in place. Obviously a secure, integrated mount with power would be used, just like the wayfarer machines have and the tablet would also be behind the drivers security screen.

cdebru

bk wrote: »

But my earlier point wasn't about the cost of the ticketing system, is was to show how ridiculously outdated and underpowered the ticketing machines used by DB are.

1 million euro in the greater scheme of things is very reasonable. Compared to the 50 million that has already been spent on Leapcard, it would be money well spent IMO.

It would fix the awful delays with leap card validation that plague the DB ticket machines and which I believe if not urgently fixed, will lead to the failure of the leap project.

It would also allow for the enabling of a great deal of functionality that the current ticketing machines can't handle.

For instance it would allow for online top-ups to be applied on buses.

It could also power many other onbus systems, such as bus location screens and CCTV. Allowing DB to save money by replacing multiple systems with jsut one system.



No, I honestly believe tablets are more durable then machines that use physical buttons. There is simply less to break. With the exception of dropping
(not a problem on a mounted bus system) I've never once seen a touch screen break on any tablet or phone. I've seen the physical buttons break (e.g. iPhone home buttons) but never a touch screen.

Again the FAA wouldn't have approved iPads for use in commercial aircraft if they weren't durable. The FAA have very strict and intensive testing of durability of all electronics used on an aircraft.

The stealing comment is just silly, obviously they aren't just going to bluetack the tablets in place. Obviously a secure, integrated mount with power would be used, just like the wayfarer machines have and the tablet would also be behind the drivers security screen.

But how much of that 50 million was spent on the ticketing hardware on bus?

I agree that the machines need to be updated but I don't think you need to reinvent the wheel there are already reliable durable machines available of the shelve that could reasonably expected to last at least 10 years with out any major reinvestment, there is nothing wrong with the current wayfarer machines other than they are not up to the requirements that should have been really obvious when they were purchased.

As has been pointed out, the use and enviroment within a cockpit has no relation to continous use for upto 20+ hours a day 7 days a week. The two don't compare.

Lastly the stolen comment is not silly I'm afraid city buses are often parked unsupervised for long periods where a wayfarer ticket machine is pretty much useless to anybody, a 400 euro tablet is an easily saleable item and drivers cabs are not that secure otherwise drivers couldn't get into them take a trip down to Marlborough st someday and see who hangs around there.

bk

cdebru wrote: »

As has been pointed out, the use and enviroment within a cockpit has no relation to continous use for upto 20+ hours a day 7 days a week. The two don't compare.

I'm sorry but it is very comparable. Look at an airline like Southwest Airlines in the US (think Ryanair), their airlines are definitely utilised 20+ hours a day 7 days a week and I can assure you the iPads are being used constantly throughout this time and have been thoroughly tested by the FAA for reliability.

It is laughable that people would think that any electronic devices on a commercial airline, in particular any as important as an electronic flight bag isn't ridiculously thoroughly tested. Way beyond anything the DB would ever do.

TheBandicoot

I'd love to get a representative from Wayfarer onto this thread to see where they see the future of on bus ticketing equipment. I'm sure they have done lots of research into the Android tablet universe. Maybe they are trying to pivot into a software business?