[leo67]

One of the cool things about my 2012 335is I'll do maybe once a month (usually with an unsuspecting passenger in the car) is to punch the accelerator at highway speeds to demonstrate the engine and transmission responsiveness and the amazing amount of torque available as the 335is overboost kicks in.

Recently performance overall has seemed a bit sluggish. It didn't have the "punch" I recall during early months of ownership (even in Sport transmission mode and with the "Sport" button on the 335is on). I dismissed it as me just getting used to the car's performance or perhaps needing to change to a different premium gasoline brand.

Today I performed the ECU reset (a.k.a adaptive transmission reset) for the first time since the 18 months I've owned the car. I took it for a test run on the Dallas North Tollway - and all I can say is WOW. The responsiveness and acceleration is vastly improved and reminds me of why I bought the car in the first place. I felt the same exhilaration as on the first test drive.

And then I got upset. Upset at how the driver adaptive algorithms apparently "neutered" my car. Upset that the method for ECU or transmission reset is not part of the owners manual. And skeptical about what the algorithms are doing and why.

Now, I'm not about to risk a libel suit. I have no malicious intent - I'm not asserting truth or fact. The driver adaptive ECU function exists in many cars, not just BMWs. My questions are of the driver adaptive ECU / transmission function in general, not singling out any particular car manufacturer or car model(s). My expressions of opinion here are based on my observations for the car I own today. I bring it to this forum in order to openly discuss and education myself and others and to solicit knowledge from technical experts with understanding of the subject matter. I'm also not a lawyer, so I hope that is enough of a disclaimer!

Is the driver adaptive ECU feature real or a myth?
Is the reset feature of the driver adaptive ECU feature real or a myth?
Is my perception of improved performance after reset only that - perception - or is it real (i.e. measurable and quantifiable)?

IF it's real, what is the reset actually doing, and what is the range of performance modifications the algorithm are applying? Or are these algorithms proprietary or trade secrets not to be revealed?

I've heard the algorithms "learn the driving style of the driver". That's of course a very qualitative description which well, basically offers no insight into what the algorithms are doing. I want to understand what my car is doing and why. My driving style varies day to day, so how can it know what I want to do when and with what characteristics? Anything that is "adaptive" takes time to learn. As I said at the first, I might do something very uncharacteristic and impulse and immediate like floor it at random times to show off my car's impressive performance (obviously, on a closed course, not on a public road). I don't think an adaptive algorithm that learned a "soft" driving style can adjust instantly to an impulsive choice to employ a "hard" driving style for a short time. (Or, maybe that's what the 335is Sport button in part does - temporary amnesia of the learned adaptations in favor of a canned, aggressive driving style? Yes, I've RTFM and the couple of lines describing what the Sport button does - but it doesn't say how or whether it interacts or overrides any learned driver adaptations.)

I've also heard the algorithms improve gas mileage. And that's where this in my mind raises a red flag. Are these algorithms reducing performance in order to improve gas mileage? Do the algorithms ever learn a "hard" driving style that actually reduces MPG (compared to a baseline MPG that would be had if the adaptive algorithms were not running at all?) Or do the algorithms only lean the one direction of trading off (reducing) performance and/or responsiveness in order to improve MPG?

Hopefully technically savvy readers can educate me and others - that's why I'm bringing this up. I'm curious, and I like to understand the nuances of my car, how I can adjust its behavior, and hear expert advice and opinion. Thanks!

As an aside, and as an entirely hypothetical scenario, if I owned a car dealership I would of course want to maximize car sales. Impressing the customer during the initial test drive whether it's a new or a preowned car would be very important. The customer would be impressed with a sports car that has great performance and also reasonable MPG. If I also had knowledge of a "secret" software feature built into the car (call it a "Performance Versus Economy Modulation Routine" that I could adjust that would maximize performance initially (particularly during the phase of the car's lifetime when test drives would be most likely to occur), I'd employ it, even if it meant the short term average the published MPG numbers may not be met. Again, we're trying to impress the customer here and get him or her to buy the car. In the longer term average over a period of many months (while the customer owns the car), the EPA numbers would be met. But in the initial months of ownership, MPG numbers may look bad - but that would be due to the car being in the break-in period when MPG suffers (as everyone knows). MPG of a newly purchased car will of course improve over time (as everyone knows). The hypothetical "Performance Versus Economy Modulation Routine" would gradually modulate (reduce) performance over a timeframe of weeks or months in order to improve MPG to meet the published EPA numbers for the car. And if the customer trades in this car years later back to me and my dealership, I would reset the "Performance Versus Economy Modulation Routine" back to maximize performance as I ready the car for customer test drives and resale as a preowned vehicle. Anyhoo, that's just my mind meandering into la-la land about a ficticious scenario and a ficticious software feature I thought up for a hypothetical car design. I certainly don't have enough money to own my own car dealership or the resources or know-how to design a car!

[F32Fleet]

#1 It's real.
#2 It's real and your actually resetting the throttle. Not the adaptations of the transmission.
#3 Change is real, but I wouldn't call it better, just different.

Throttle response has nothing to do with MPG. It's all about the driver.

[leo67]

Thanks for the info.

This makes sense, and I'm glad you have the definitive answer about what the reset is actually resetting (throttle mapping not transmission).

Many mistakenly call this a "adaptive transmission reset". This could be because 1) those who perform the reset perceive a change in transmission behavior and/or 2) because the adaptive transmission management feature is documented by BMW and described to learn a driver's style (which anecdotally is the culprit we are trying to "undo" by a reset).

The adaptive transmission management (ATM) function is documented as a BMW technology that "learns the driver's style" and "this pattern is continuously monitored, so that ATM always adapts to the style of the current driver."
"Current driver" implies to me that the ATM function has no long term memory about previous drivers or how you drove the car a few days weeks or months ago. So even if you could reset the driver adaptive values in ATM, I suspect it wouldn't have a noticable effect. http://www.bmw.com/com/en/insights/t...anagement.html.

However, I would guess that the changes enabled by the throttle reset do persist (how long I don't know) and I also suspect that it could indirectly affect transmission behavior (which gears are selected and when), which is what many perceive.

Throttle mapping changes could cause the driver to alter accelerator pedal input. Accelerator pedal and engine data are inputs to the adaptive transmission management function. Change the input values to a subsystem (transmission), and you could alter its behavior or its outputs. If it does and to what extent would require detailed knowledge of ATM and/or empirical data. Or, accept the qualitative descriptions of the effect (I personally would say my car was "more responsive". After the reset I perceived better engine and transmission response, and better acceleration at highway speeds). Or I could be imagining the improvement - that's a possibility too!

At some point the throttle and transmission algorithms adapt and revert to something close the prior state before the reset. How long? days? weeks? If I'm convinced this is having a beneficial effect, then it'd be time to perform a throttle reset again.

[Artmasterx]

FYI, I think that "current driver" is meant to imply the "current key" being used. That is, if a husband and wife each have their own key, the car recognizes each key separately and stores a separate adaption for each key.

So it has a long-term memory for each key but can instantly switches the adaption based on which key/person is operating the vehicle.

[leo67]

I've seen posts referring to these driving styles being stored in the particular key (remote control) being used. And reserving one key for "Miss Daisy" driving and another key for "Mario Andretti" driving. It's not listed in the owner's manual as part of the Profile data that's stored in the key (it might be stored as part of the Service data, but is is loaded and used by the car)? Admittedly I wouldn't expect something the car is doing "behind the scenes" to be part of the Owner's manual which focuses more on the visible, tangible driver/car interfaces. But the curious Owner would like to know whether the key contents are loaded which affect driving characteristic, or if this is just speculation.

Lacking technical documentation to verify, someone with these two keys could compare driving characteristic with one key versus the other over an identical route. Make it a "blind" comparison (not knowing which key was used until the conclusion of both tests and after recording data/assessment of each to avoid introduction of any "placebo" bias).