Common ECU + Software

Discussion and debate about the MotoGP class
Post Reply
User avatar
Fingernails
Posts: 1584
Joined: Mon Sep 22, 2014 1:41 pm

Common ECU + Software

Post by Fingernails »

Hi all. First time poster.

Is there anywhere I could find out more information about the common ECU and shared software that is coming bit by bit to MotoGP.

In particular, I'm confused by the common software. If it was a common ECU with bespoke software for each team, then that still allows teams to develop all sorts of private technology in their own software. But, if the software itself is common, then surely for any team to make sure of new technology, they'd need to add it to the software, in which case the technology would be revealed to other teams, and they could make use of it. If so, that's a significant levelling of the playing field. However, I've read that some software functions suitable for the Ducati has been added to the current software, but that some other 'open' teams do not have the technological wherewithall to make use of it. Does that mean that it would be possible for (e.g.) Honda to add functions to the software giving them a technological advantage, but where the use of these functions was obscured so that other teams could not understand nor make use of them?

How does/will/can it all work?

Also, is this new standardisation a reason why there has been an influx of new factory teams, Aprilla and KTM? (I don't count Suzuki here as they appeared to be headed back to MotoGP way before the ECU rules for the future were knwn).

User avatar
Kropotkin
Site Admin
Posts: 3160
Joined: Tue May 20, 2008 9:44 pm
Location: Dieren, The Netherlands
Contact:

Re: Common ECU + Software

Post by Kropotkin »

There's an interview with Corrado Cecchinelli, Director of Technology, on the main page

http://motomatters.com/interview/2014/0 ... oal_o.html

That explains a lot of it.

The process is that from 30th June 2015, the website will be open to all manufacturers to contribute functionality requests and functionality. Everyone with a log in will be able to see those requests, so yes, the manufacturers will be sharing technology. That means they won't be giving away their secrets, and that will keep the level of the electronics relatively limited.

In theory, Honda or Yamaha could add very complex electronic functions which only the factory teams could use, but Cecchinelli has the final say on what goes in or out, and if they try anything like that, their request will be rejected. The aim is to make the software user friendly, for a sophisticated user.
--
http://www.motomatters.com/ - MotoGP News, Analysis and Race Reports

User avatar
Fingernails
Posts: 1584
Joined: Mon Sep 22, 2014 1:41 pm

Re: Common ECU + Software

Post by Fingernails »

Thanks. That answers some questions, particularly that the software functions must be usable for any riders. There must be the potential for bike and engine specific data, such as engine mappings. But that sort of detail must already have been sorted for the Open class this year.

Disappointing in some ways that the software won't be fully open source. I could imagine software savvy fans going through the code to see what they can find in it

andrebt
Posts: 549
Joined: Wed Nov 06, 2013 6:57 am
Location: orlando, florida

Re: Common ECU + Software

Post by andrebt »

To be honest initially I was skeptical about whole the common software/electronics deal(too many unknowns when it come to these things) , but seeing how the smaller teams stand to gain something it should be interesting.

User avatar
Fingernails
Posts: 1584
Joined: Mon Sep 22, 2014 1:41 pm

Re: Common ECU + Software

Post by Fingernails »

andrebt wrote:To be honest initially I was skeptical about whole the common software/electronics deal(too many unknowns when it come to these things) , but seeing how the smaller teams stand to gain something it should be interesting.
I find it difficult to imagine how the software could work for all bikes. Surely the software would read information from all sorts of sensors and control all sorts of engine and other parameters. These combinations would surely be bespoke for each factory bike, and the question is what happens when they are all put together into one piece of software. It seems that Ducati has added a lot of functionality into the software, and presumably this is to enable what Ducati wants to do with their bike. When Honda and Yamaha come to add their desired bits to the software, will they be able to add in their current software to control (e.g.) their seamless gearbox the way that they want to, or will they have to adapt their bike's technology to the software?

Looking at two possible extremes of how the software might work, it could be that the ECU is pretty much hard-wired to run the a generic motogp bike, and everyone who uses it has to engineer their bike so that it presents itself to the ECU as that generic bike. That could strongly limit what technology could be added to the bikes. E.g. a seamless gearbox might have to work in one particular way. That would very much level the playing field between manufacturers. And would be a big plus for Dorma because you couldn't get a strong team dominating the championship and providing too many customer bikes, so that this team had too much bargaining power.

Or, the ECU software might be very general, with the ECU itself having a good amount of general I/O channels, and a plugin architecture for the software whereby different plugins can be loaded by different teams. That might even allow a team to add high technology functions to the software without revealing their technology. In that the software might read information from input pins, and send signals out of output pins. Without information as to what sensors these pins receive information from and what they control on the bike, that could make it difficult to work out what that technology does. Even more so if (e.g.) a team doesn't need to use all the I/O pins on the ECU, and hence could read and send dummy information to/from unused pins to hide what the software is truly doing.

That last paragraph is of course an unrealistic extreme scenario. As Corrado Cecchinelli has said, software updates won't be added to the software unless he's confident that they can be used by reasonably competent smaller teams. So anything exhibiting any obfuscation should be simply rejected. However, I would guess that the teams will bring in the best brains they have to work out how to get the best technology they can into the software with the minimum spread of that technology to other teams. As we can see with hacking, software people can be remarkably ingenuous. I would expect that teams will consider adding Trojan Horse functions into the software, though whether or not there would be an actual attempt is another question. And of course whether or not any such attempt would be successful is another question again.

In any case, the system will need data. It's mentioned in the interview that turn-by-turn suspension settings require a fair sized team of people to enter that data for each track, something that smaller teams cannot afford. Even if the software is equalised, track and other data may not be, still allowing the teams who can spend more a significant advantage. Apart from the mentions of turn-by-turn suspension stiffness, there must be a large amount of other such data, such as engine mappings. It would again be the advantage of the big teams to add functions that require large amounts of data input, as smaller teams wouldn't be able to afford the data creation and entry. Will data be shared among teams? In computing there is sometimes a grey area between what is program code, and what is data. (E.g. Logic or Functional Programming, or a rule based system) If data doesn't have to be shared, only code, then this potentially opens a loophole for secret code functions to be moved to the data sphere, rather than the code sphere which must be shared. I would guess that the teams would police each other to a certain extent, as they would see all updates, and could use their own clever people to scan updates to see if any loopholes have been exploited, or anything proposed that is against the spirit of the rules. If that happens in practice, then of course there's the risk of teams complaining and trying to block anything that they themselves can't or don't want to benefit from. Meaning that it could become difficult for overseers to tell the difference between objections for valid reasons, or knee-jerk objections.

There might be a competitive advantage for the big teams to come up with other hardware on their bikes, e.g. technologies for seamless gearboxes, that are expensive to develop and very particular. Therefore even if other teams knew how the software worked, they still wouldn't be able to apply it as they wouldn't have the mechanical IP to build the mechanical parts of the system. That may become a driver for greater concentration on the mechanical parts of the bike and engine. Or it could be that these systems are already near optimal, meaning that there is limited scope for a team to build a technical advantage in that manner.

It is true that the open software is already in use. But, there is only one large well resourced team using this software. Hence it's possible that even if there have been no problems so far, that things could get much more difficult to manage once all the factory teams are using the same software. There are comments in the interview about smaller teams not being able to use some of the Ducati updates. Knowing exactly what it was about those updates that made it difficult to use (e.g. excessive data creation requirements, obfuscation, who knows what) would be very revealing.

I like to think of myself as not being particularly paranoid. However, Corrado Cecchinelli has strong past links with the Ducati team. I have no reason to believe that he'd be biased towards one team or another (Jean Todt in F1 does not seem to have favoured Ferrari for example), but there could be a risk of conspiracy theories growing concerning that link. Whether conspiracy theories are of any concern or not, I don't know. Personally I wouldn't be concerned about this unless there was strong evidence that anything was going on. But, as they're currently the underdog I would like to see Ducati return to the front and win championships again. It would dilute that enjoyment somewhat if there were accusations of favouritism because a 'Ducati man' was in charge of the software.

User avatar
JanBros
Posts: 1887
Joined: Thu Mar 11, 2010 7:57 pm
Location: The land of Francorchamps

Re: Common ECU + Software

Post by JanBros »

wow, what's that all about :shock:

first : there is no factory team yet (Ducati) that runs the open software. They were gonna, but last minute changes to the rulebook meant they are running their own software.

second : there won't be any possibility of programming parts of the software (that is what I believe) by the teams. Because if that was possible, it wouldn't be the same software. you have to look at it like it's a very powerfull power commander (or that is what it should be) , like for example when :
- throttle = 50%
- engine temp = X°
- rpm = XX.000
- lean angle = Y°
- P mm suspension travel
-... other inputs

the teams can select the outputs, like :
- Z° advance ignition
- Q% of mixture richness
- ...

and they can probably choose from 0 to 100% for each output.
Bi-stroker Parallelus Bi-cylindricus

User avatar
Fingernails
Posts: 1584
Joined: Mon Sep 22, 2014 1:41 pm

Re: Common ECU + Software

Post by Fingernails »

JanBros wrote:wow, what's that all about :shock:

first : there is no factory team yet (Ducati) that runs the open software. They were gonna, but last minute changes to the rulebook meant they are running their own software.
OK, but there have been mentions of an update to the Open software 'inspired by Ducati'. Even if they aren't running the software in the Factory 2 class, there still appears to have been an update that the smaller teams cannot use. http://www.crash.net/motogp/news/201007 ... pment.html http://www.crash.net/motogp/news/200820 ... ision.html And if Ducati aren't yet running the Open software, that means that there is not yet any experience of running this collaborative structure (currently a mailing list) with a Factory involved, which means that the unknowns that I am concerned about are even more unknown.
second : there won't be any possibility of programming parts of the software (that is what I believe) by the teams. Because if that was possible, it wouldn't be the same software.
The first paragraph of the article with the interview says that teams will be able to submit code in the new collaborative environment. (My emphasis)

http://www.motomatters.com/interview/20 ... oal_o.html
From 2016, the entire MotoGP class will switch to a single, spec software for the electronics on the bikes. Development of the software is to become a collaborative process, with the factories competing in MotoGP supplying code and requirements through a single website. This much we know. But what we don't know is much more interesting. Which technologies will be supported? Which functions will be available? How sophisticated will the software be? Who will lead the software process, the factories or Dorna?
Hence, the open software will include code written by teams. Which is the primary cause of the things I discuss in my most recent post being plausible.
you have to look at it like it's a very powerfull power commander (or that is what it should be) , like for example when :
- throttle = 50%
- engine temp = X°
- rpm = XX.000
- lean angle = Y°
- P mm suspension travel
-... other inputs

the teams can select the outputs, like :
- Z° advance ignition
- Q% of mixture richness
- ...

and they can probably choose from 0 to 100% for each output.
What you have there is a rule. All that is needed is for rule to be able to set internal state variables, the ability to write multiple rules, and you'll have the power of a full Turing Machine.. See Rule based system. If the rules can't set internal state variables, then the amount of programming possible by creating a rule set will be limited, but still an awful lot will be possible through the rules. There would just have to be a lot of them. I don't know if the open software allows state variables, particularly the creation of them, but most control systems (which the ECU is) would use them. Hence, your example doesn't argue against the possibility of teams being able to write their 'code' by effectively moving their algorithms from the agreed and shared code into non-shared data. This doesn't mean that the current system will allow it, but it's still plausible at this time.

andrebt
Posts: 549
Joined: Wed Nov 06, 2013 6:57 am
Location: orlando, florida

Re: Common ECU + Software

Post by andrebt »

Fingernails wrote:
andrebt wrote:To be honest initially I was skeptical about whole the common software/electronics deal(too many unknowns when it come to these things) , but seeing how the smaller teams stand to gain something it should be interesting.
I find it difficult to imagine how the software could work for all bikes. Surely the software would read information from all sorts of sensors and control all sorts of engine and other parameters. These combinations would surely be bespoke for each factory bike, and the question is what happens when they are all put together into one piece of software. It seems that Ducati has added a lot of functionality into the software, and presumably this is to enable what Ducati wants to do with their bike. When Honda and Yamaha come to add their desired bits to the software, will they be able to add in their current software to control (e.g.) their seamless gearbox the way that they want to, or will they have to adapt their bike's technology to the software?

Looking at two possible extremes of how the software might work, it could be that the ECU is pretty much hard-wired to run the a generic motogp bike, and everyone who uses it has to engineer their bike so that it presents itself to the ECU as that generic bike. That could strongly limit what technology could be added to the bikes. E.g. a seamless gearbox might have to work in one particular way. That would very much level the playing field between manufacturers. And would be a big plus for Dorma because you couldn't get a strong team dominating the championship and providing too many customer bikes, so that this team had too much bargaining power.

Or, the ECU software might be very general, with the ECU itself having a good amount of general I/O channels, and a plugin architecture for the software whereby different plugins can be loaded by different teams. That might even allow a team to add high technology functions to the software without revealing their technology. In that the software might read information from input pins, and send signals out of output pins. Without information as to what sensors these pins receive information from and what they control on the bike, that could make it difficult to work out what that technology does. Even more so if (e.g.) a team doesn't need to use all the I/O pins on the ECU, and hence could read and send dummy information to/from unused pins to hide what the software is truly doing.

That last paragraph is of course an unrealistic extreme scenario. As Corrado Cecchinelli has said, software updates won't be added to the software unless he's confident that they can be used by reasonably competent smaller teams. So anything exhibiting any obfuscation should be simply rejected. However, I would guess that the teams will bring in the best brains they have to work out how to get the best technology they can into the software with the minimum spread of that technology to other teams. As we can see with hacking, software people can be remarkably ingenuous. I would expect that teams will consider adding Trojan Horse functions into the software, though whether or not there would be an actual attempt is another question. And of course whether or not any such attempt would be successful is another question again.

In any case, the system will need data. It's mentioned in the interview that turn-by-turn suspension settings require a fair sized team of people to enter that data for each track, something that smaller teams cannot afford. Even if the software is equalised, track and other data may not be, still allowing the teams who can spend more a significant advantage. Apart from the mentions of turn-by-turn suspension stiffness, there must be a large amount of other such data, such as engine mappings. It would again be the advantage of the big teams to add functions that require large amounts of data input, as smaller teams wouldn't be able to afford the data creation and entry. Will data be shared among teams? In computing there is sometimes a grey area between what is program code, and what is data. (E.g. Logic or Functional Programming, or a rule based system) If data doesn't have to be shared, only code, then this potentially opens a loophole for secret code functions to be moved to the data sphere, rather than the code sphere which must be shared. I would guess that the teams would police each other to a certain extent, as they would see all updates, and could use their own clever people to scan updates to see if any loopholes have been exploited, or anything proposed that is against the spirit of the rules. If that happens in practice, then of course there's the risk of teams complaining and trying to block anything that they themselves can't or don't want to benefit from. Meaning that it could become difficult for overseers to tell the difference between objections for valid reasons, or knee-jerk objections.

There might be a competitive advantage for the big teams to come up with other hardware on their bikes, e.g. technologies for seamless gearboxes, that are expensive to develop and very particular. Therefore even if other teams knew how the software worked, they still wouldn't be able to apply it as they wouldn't have the mechanical IP to build the mechanical parts of the system. That may become a driver for greater concentration on the mechanical parts of the bike and engine. Or it could be that these systems are already near optimal, meaning that there is limited scope for a team to build a technical advantage in that manner.

It is true that the open software is already in use. But, there is only one large well resourced team using this software. Hence it's possible that even if there have been no problems so far, that things could get much more difficult to manage once all the factory teams are using the same software. There are comments in the interview about smaller teams not being able to use some of the Ducati updates. Knowing exactly what it was about those updates that made it difficult to use (e.g. excessive data creation requirements, obfuscation, who knows what) would be very revealing.

I like to think of myself as not being particularly paranoid. However, Corrado Cecchinelli has strong past links with the Ducati team. I have no reason to believe that he'd be biased towards one team or another (Jean Todt in F1 does not seem to have favoured Ferrari for example), but there could be a risk of conspiracy theories growing concerning that link. Whether conspiracy theories are of any concern or not, I don't know. Personally I wouldn't be concerned about this unless there was strong evidence that anything was going on. But, as they're currently the underdog I would like to see Ducati return to the front and win championships again. It would dilute that enjoyment somewhat if there were accusations of favouritism because a 'Ducati man' was in charge of the software.
even though the article main pertains to cost saving, Toby Moody pretty seem to have an idea of how the smaller teams stand to benefit financially from from the move to a common ecu/software package:

http://plus.autosport.com/free/feature/ ... he-future/

smAshmoto
Posts: 670
Joined: Mon Sep 02, 2013 4:01 am
Location: Adelaide, Australia

Re: Common ECU + Software

Post by smAshmoto »

This whole thing makes my head hurt.

I just want to watch motorcycles go fast.
"Troubled childhood? If you call being a 9 year old kid with a 35 year old girlfriend troubled."

-Chaz Michael Michaels, ice devouring sex tornado

User avatar
Fingernails
Posts: 1584
Joined: Mon Sep 22, 2014 1:41 pm

Re: Common ECU + Software

Post by Fingernails »

smAshmoto wrote:This whole thing makes my head hurt.

I just want to watch motorcycles go fast.
I'm sure that you would also like to see a number of teams competing on an equal footing at the front, rather than just one team build up an unassailable technological advantage and dominate the series for the foreseeable future. Without some equalisation of software, we were rapidly heading in the direction of the latter. IMHO having open software is one way of equalising the field, provided that it's done properly and loopholes are not left open.

User avatar
Tourn46
Posts: 3402
Joined: Sun May 15, 2011 5:09 pm
Location: UK

Re: Common ECU + Software

Post by Tourn46 »

Fingernails wrote:
smAshmoto wrote:This whole thing makes my head hurt.

I just want to watch motorcycles go fast.
I'm sure that you would also like to see a number of teams competing on an equal footing at the front, rather than just one team build up an unassailable technological advantage and dominate the series for the foreseeable future. Without some equalisation of software, we were rapidly heading in the direction of the latter. IMHO having open software is one way of equalising the field, provided that it's done properly and loopholes are not left open.
Instead, we will see the Open/Sattelite/Production/CRT/'Whatever they become in the future' Honda's, Yamaha's and Ducati's mechanically kept just that bit slower than their factory counterparts.

User avatar
MiniNinjaMk5
Posts: 1713
Joined: Tue May 08, 2012 12:57 am

Re: Common ECU + Software

Post by MiniNinjaMk5 »

That's always been the case though Tourn. But at least this will be one less advantage that the factory teams have and the non-factory teams might be able to get a bit closer to the factory guys.

Not sure if it will do enough to give the first non-factory win since 2006 but it's a step towards a more level playing field, like the move to one-tyre suppliers was before it.

User avatar
Tourn46
Posts: 3402
Joined: Sun May 15, 2011 5:09 pm
Location: UK

Re: Common ECU + Software

Post by Tourn46 »

MiniNinjaMk5 wrote:That's always been the case though Tourn. But at least this will be one less advantage that the factory teams have and the non-factory teams might be able to get a bit closer to the factory guys.
Yeah, I won't disagree with that.

User avatar
Fingernails
Posts: 1584
Joined: Mon Sep 22, 2014 1:41 pm

Re: Common ECU + Software

Post by Fingernails »

MiniNinjaMk5 wrote:That's always been the case though Tourn. But at least this will be one less advantage that the factory teams have and the non-factory teams might be able to get a bit closer to the factory guys.

Not sure if it will do enough to give the first non-factory win since 2006 but it's a step towards a more level playing field, like the move to one-tyre suppliers was before it.
I do like the idea that technology is shared among teams, at least to a small degree. Even among different factory teams. If that means that certain technologies won't be put in the racers because the teams don't want to release the technology to others. But, isn't the whole patent system meant to allow disclosure of inventions while still protecting the owners of those inventions? Can patented technology be put into the Open Software without losing patent protection?

Post Reply