One of the ways in which MotoGP has attempted to control both cost and performance has been through the use of spec electronics. The first step was to make the ECU, the computer hardware, standard, allowing factories to continue to run their own software on the spec Magneti Marelli ECU adopted in 2014. This move prevented factories from developing their own specialized hardware and leveled ECU performance.
In 2016, MotoGP switched to spec software on top of the spec hardware. With everyone forced to use the same, standardized software, factories could no longer throw large numbers of software engineers at the problem to try to figure out more elegant and efficient ways of control the behavior of the bike, through traction control, engine braking, and anti-wheelie strategies. Dorna had hoped to create a level playing field with this move.
Of course, there is nothing engineers love more than challenge of finding ways to tilt a level playing field in their favor. Since the adoption of spec software, the different factories have find different ways of trying to extract an advantage from the current rules.
One of the avenues engineers have explored is the use of an IMU, an inertial measurement unit. What an IMU does is report the lean angle, attitude, and acceleration of a motorcycle. By its very nature, it requires a lot of intelligence to measure and calculate all of these factors. And the suspicion has arisen that the factories are taking advantage of that intelligence to use the IMU as a sort of secondary ECU, capable of doing more calculations and modifying the inputs to the spec ECU to change the behavior of the bike. The IMU is a so-called 'free sensor', meaning that factories are free to choose which IMU they wish to use.
Enter the spec IMU
This is about to end. From 2019, Dorna will supply a spec IMU, standardized and controlled to prevent factories from exploiting it. With the software and processing power of the IMU limited, factories will no longer be able to subtly alter the behavior of the spec ECU in pursuit of an advantage.
At Le Mans, I spoke to Corrado Cecchinelli, Director of Technology for MotoGP, about why Dorna had decided to go to a spec IMU. "For two reasons, like always," he explained. "The main reason this time is security or anti-cheating or anti-tampering, call it whatever. Because the IMU is a very important sensor that by nature includes a chip and some logic inside, so you cannot say, "you cannot put logic inside your IMU". So once you have to allow logic, you have to allow everything you can control. This is the main reason. Second reason is as always cost controlling, because as this is a free sensor everybody is trying to develop his own, which is a lot of wasted money from our perspective."
There have been suspicions around the paddock that some factories had been using the IMU as a sort of "piggyback" ECU, with the IMU taking some of the calculation load from the spec ECU, and allowing more precise control of the bike behavior. The electrical system of MotoGP bikes, like most modern vehicles, is connected using a so-called CAN bus, which effectively operates as a miniature network aboard the vehicle. All of the output of the various sensors gets sent over the CAN bus, and the ECU takes this output and makes decisions on controlling the power delivery accordingly. But as the CAN bus is an open network, all of the sensors connected to the CAN bus can theoretically read data from it, and do their own calculations based on that data. An intelligent IMU could read data from the CAN bus, then modify its own output based on that input, allowing more precise control of the ECU software strategies.
This is part of a semi-regular series of insights into the world of motorcycle racing, exclusive for MotoMatters.com site supporters. The series includes background information, in-depth analysis, and opinion pieces. Though the majority of content on MotoMatters.com remains free to read, a select amount of the more interesting content will be made available solely to those who have supported the website financially by taking out a subscription.
The aim is to provide additional value for our growing band of site supporters, providing extra original and exclusive content. If you would like to read more of our exclusive content and help MotoMatters.com to grow and improve, you can join the growing band of site supporters, by taking out a subscription here.
Comments
When I saw the spec rules
When I saw the spec rules come out a couple of years ago, but saw the clause on the ICUs, I said to myself, "Uh oh". I wonder if this is Yamaha's problem, they lag behind in enhanced rule bending.
In reply to When I saw the spec rules by mtiberio
hmm
not sure if that explains Zarco's amazing performance.
Penalties
So if and it's a big if, some factories are doing this and cheating will there be disqualifications? Would Dorna take action or brush under the carpet.
In reply to Penalties by Ppparkinson9
Letter vs spirit.
The question is how would you prove it was cheating. Or to use a more Formula 1 inspired train of thought, where in the rules does it say that the external sensors have to provide 'true' data to the spec ECU? From next year the answer is obvious, the spec sensors have their capabilities and that is that, any tampering is banned. At the moment the IMU is a free design, it can do whatever the manufacturer wants it to do, if that means it feeds the spec ECU false data to manipulate the bike's behaviour then so be it, the ECU itself hasn't been altered nor has it's software, so no rules are broken.
.
It's a genius way to interpret the rules, it might not be very fair, but as the old saying goes, fair fights are for suckers.
In reply to Penalties by Ppparkinson9
It would not be cheating
the rules specifically allow everything mentioned above, they even include a wiring diagram showing exactly how the secon IMI is to be piggy backed into the ECU.
Now from 2019 onwards then that would be a different matter as then it would not be allowed anymore.
I can see the job postings now
Data Encryption Specialist
Job description: Join a dynamic, fast paced team working with big data in real time. Your role would be to work on encryption and decryption of data, and interpretation of data + parsing of commands over the CANBUS protocol.
Job requirements:
I see 5 of these ads going up as Yamaha thinks it can decrypt the ECU over a lunch break.
In reply to I can see the job postings now by CTK
Encryption
There are plenty of off the shelf encryption algorithms that can be used to sign software. If someone can figure out how to break these then they'd have a much more lucritive career in bank fraud than in ECU programming.
I wonder if we shall see the
I wonder if we shall see the effect of this change in 2019 when, say for example Honda, all of a sudden start having random crashes or a slower lap time that isn't specifically explanable via the normal riding/racing rules. Or if Yamaha for example become a lot better with no lap time changes for themselves.
Any more scenarios like increase TC with more lean?
Quite technical write up. Thanks a lot. Can some one share some other scenarios like this? Very interesting to know.
On side note, I think team yamaha will go back 5 years now with already struggling ECU adaptation woes, and now plus spec IMU.
In reply to Any more scenarios like increase TC with more lean? by aj700
More scenarios
There are a lot of scenarios you can imagine. The spec ECU is relatively limited, and can't do corner by corner TC. So the IMU could alter the lean angle output for each corner to achieve the level of TC required. It could add or subtract acceleration data (the IMU also outputs acceleration force) to give more or less wheelie / TC / wheel spin out of particular corners. In a corner with a lot of camber, it could add or subtract lean angle to allow for more or less engine braking, for example. The possibilities are not hard to imagine. They are just a Small Matter of Programming.
Simple Example for BMWF800 ECU
For the BMW F800 series, someone sells a temperature sensor with a chip which adjusts the volts/current ratio to the "intake air input" for the ECU. This tells the ECU that the air temperature is colder than it actually is, thus causing the ECU to allow more fuel, running the engine richer. The standard ECU runs the air/fuel ratio quite lean for fuel economy/emissions, so this cheaply gives more power without buying a Power commander type aftermarket ECU and having to work out your own maps.
The richest factories will still hire the smartest engineers !
In reply to Simple Example for BMWF800 ECU by als02
Too smart
A similar example for my KTM690. The lambda sensor doesn't work properly with an aftermarket exhaust and custom mapping, and when it's out of range it just reports zeros, which the ECU interprets as an error. There's a little box that you can fit which just sends a random results to the ECU. With the custom map the ECU isn't using the lambda data anyway, so you just want to stop the ECU from reading an error.
Yamaha-
Yamaha-
Their issue isn't substantially related to this. They have been attempting to establish their own understanding of the Magneti Marelli electronics. Ducati had been using it already, to very good effect, and had a leg up. Honda hired a number of key staff away from MM to work on their project. Rumor has it Yamaha have decided to go ahead and hire a MM specialist too.
Yamaha also relied very heavily on their advanced proprietary package to keep tires holding up while at lean more. Honda's front smash then turn is different. Yamaha had more to sort and chose to do so w less MM help. The tune may be changing now re how they are working on it, but they still have a tough go of it. Plus Marc Marquez, he helps. Plus losing Zarco because Yamaha has been lacking some basic smarts of late, doesn't help for next year.
I doubt we will find out that this has been a huge deal relative to other factors.
In reply to Yamaha- by Motoshrink
Magnetti not new to Yamaha
Yamaha was also on magneti electronics like ducati prior to the spec hardware and software introductions.
Honda and Suzuk were not, using their own in hose delveloped componets. For Honda it took a compete rethink of their entire controls. They played with predictive algorithms to control based on equations and torque output sensor at shaft. They had the most to learn and therfor have the momentum behind them verus the stagnant yahama who had the least to change and learn.
Suzuki benefited the most. Able to leave behind a substandard system that otherwise corporate pressures to keep what they developed would not alow, having been forced to change was a big plus that otherwise they could not convince the company to leave behind.
Cue...
more intelligent wheel speed sensors...
until ALL sensors are spec, and the data path between sensor and spec ecu is encrypted (to prevent the team building an interceptor box to tweak the signal in flight) the fun and games will continue.