A Mechanic's Life - Tales From Under the Hood

roadburner

Mr_Shiftright said:

My tax accountant claims that TurboTax and other tax software people use are directly responsible for his brand new Lexus and his second home in Scottsdale AZ.

I used to prepare taxes- it was something attorneys were expected to do in rural communities at that time. I used CCH's Computax system. When I left the practice I prepared my own taxes myself. I finally switched to Turbo Tax when I was convinced that it came up with the same numbers I did. That said, I switched to a tax accountant a few years ago when my returns became more complex.

I don't think TurboTax is inaccurate; the problem is that the taxpayer doesn't know what he/she doesn't know. Garbage in, garbage out....

roadburner

@stever: Here is the BMW Fx coding discussion that I mentioned previously. The number of parameters that can be modified is amazing. I will be so happy when I'm able to deactivate the idiot warnings and disclaimers.

Mr_Shiftright

I'd like to learn how to do that--but I'll wait until you do it first

Mr_Shiftright

RB--you are exactly right about the TurboTax comment--people get in over their heads.

Reminds me of some people who get into do-it-yourself mods on their cars. They don't have a rational plan and don't connect the dots, and then end up with FrankenCar.

thecardoc3

stever said:

Sure, there's tons of free or cheap software out there that's as good or better as the stuff you can buy at the app store.

There are just as many of those that are Trojans that look real nice but are doing all kinds of nasty things in the background. If you want to risk infecting your PC or smart phone go right ahead. But risking someone else's health and safety isn't a right and that's what open sourcing for the software that is running in the cars amounts to.

thecardoc3 said:

Say's the guy that can't identify the right oil for a given car let alone advise an owner how to.

stever said:

You lost me here, I know what oil my car uses and I know how to tell people to look it up in the manual for their car.

So are you saying that you do finally know the difference between a brand and a specification?

roadburner

Mr_Shiftright said:

Reminds me of some people who get into do-it-yourself mods on their cars. They don't have a rational plan and don't connect the dots, and then end up with FrankenCar.

I learned that lesson when I modified my first small block Chevrolet. If I had it to do over again I would still have hot rodded it, but I would have gone with a milder cam, not installed headers, and learned how to properly calibrate a Rochester Quadrajet. I'm about due a request to rebuild one....

roadburner

Mr_Shiftright said:

I'd like to learn how to do that--but I'll wait until you do it first

The more I read the more I'm tempted; I don't want to make any major tweaks- just shut off the seatbelt gong post-startup, remove the legal disclaimer from the iDrive screen, and add the ///M splash screen.
For now, anyway...

stever

thecardoc3 said:

So are you saying that you do finally know the difference between a brand and a specification?

My radar is always attuned to "marketing".

Virusware is sort of a red-herring. You (or perhaps I should say, your customers) may be more likely to be hit by a trojan that the freeware outfits. Juicier target.

thecardoc3

Once some people start down a path there is nothing to stop them from going places that they shouldn't.
This link. http://thehackernews.com/2015/10/hacking-car-airbag.html is just one of many where Trojan software not only lets people play with their own cars, it creates a loophole that infects the professional tools and even the shop's computer systems should a shop ever have to work on the car in the future.

Consider. We have cars today that can apply the brakes if the driver fails to brake soon enough. Who are the lawyers going to chase if a vehicle get's infected and that system fails to operate or operates too late to be effective?

stever

roadburner said:

Mr_Shiftright said:

I'd like to learn how to do that--but I'll wait until you do it first

The more I read the more I'm tempted; I don't want to make any major tweaks- just shut off the seatbelt gong post-startup, remove the legal disclaimer from the iDrive screen, and add the ///M splash screen.
For now, anyway...

Just watch for that Code Default Values setting LOL.

stever

thecardoc3 said:

Consider. We have cars today that can apply the brakes if the driver fails to brake soon enough. Who are the lawyers going to chase if a vehicle get's infected and that system fails to operate or operates too late to be effective?

Now we're getting to the crux and it's about time. Code tweaking is already a fait accompli and that cat is going to be hard to stuff back into the box.

thecardoc3

stever said:

Just watch for that Code Default Values setting LOL.

This could fix that.

https://www.youtube.com/watch?v=qGvyhQJA4EI

stever

lol, I could have used that roof rack. Where's the part where they make a coffee table out of the bits?

thecardoc3

Finally someone that gets it. http://www.autonews.com/article/20160215/RETAIL05/302159972/putting-the-back-shop-front-and-center

"We have an industry that treats technicians like second-class citizens, and it drives me nuts. And I'm going to do everything I can for the rest of my career to change that."

thecardoc3

Here is another article worth reading. http://www.autonews.com/article/20160215/RETAIL05/302159974/the-world-according-to-mark

The tendency to "crucify" technicians for mistakes

"How long has he been here? What level of training has he had?" We ask all those questions and then we go: "OK. He took off 97 bolts, pulled the motor, waited three weeks just for the parts to get here -- and had to remember how to put it back together." Give him a break and move on. He's a good guy.

The trade has no tolerance for mistakes and of course that's because the customers don't have any tolerance for them. Everybody makes mistakes.

Compensation

It's easy for a good tech today to make one-hundred grand a year. Put that in The Wall Street Journal and hopefully 30 years from now we'll see another generation of guys who want to work on cars and enjoy life.


It would be great of this was true, and maybe it is in some areas where the cost of living drives the wages but a quick look at the Bureau of Labor Statistics reveals the truth. http://www.bls.gov/oes/current/oes493023.htm

Locker rooms
Walk into a technician locker room in a dealership and you'll know what the dealer thinks about his technicians.


That's assuming they actually do think about the technicians.

Shop environment

We need bright people to work on cars today. Then we institutionalize them in this 1940s sweat-shop atmosphere.


The first step towards fixing a problem is realizing exactly what the problem is.

stever

Broken system and having nice locker rooms is just a band aid.

Mr_Shiftright

Well after a few weeks of crap jobs at the dealership and getting beat up by the flat rate, and thereby be surrounding by equally disgruntled techs who sit around and complain all day, I'd want the hell out of there.

Really, the tech schools do not train the technician for the Total Reality of working on modern cars. They don't teach basic money management and the ins and outs of debt (for tools, equipment, etc) , how to relate to customers (or his own service writer), and of course, they cannot teach him what only experience can teach him.

thecardoc3

How about one more piece of the puzzle. Remember all of the "sting" stories? Why did the dealers rarely (if ever) have the tech's backs and call the reporters on any parts of their stories that were inaccurate, if not just plain wrong?

The two most likely reasons are; They were either complicit to the story being written to some degree, or else they really didn't care beyond the token gesture of replacing any of the employee's that were targeted. The really sad part of all of that was how the dealer would usually get excused as all the blame was directed to the employee's while the policies at the dealer were set up so that they would use rewards and punishments to steer the next employee right back into doing what got the last one in trouble.

Mr_Shiftright

Oh I don't know...some of these "sting" stories do allow the shop to defend themselves, and quite frankly, some of the "stings" are pretty slam-dunk fraud.

But yes, there are cases where the "sting" is pure BS.

thecardoc3

Mr_Shiftright said:

Oh I don't know...some of these "sting" stories do allow the shop to defend themselves,

When they do, they try to portrait it as the dealer being the victim of bad employee's

Mr_Shiftright said:

and quite frankly, some of the "stings" are pretty slam-dunk fraud.

Agreed, just don't give the management a free pass.

thecardoc3

Here is something that you can use to see how much you understand computer controls. The vehicle is a 1997 Jeep Wrangler with a 2.5l, five speed manual transmission. The owner has modified it by adding a turbo charger. It bucks and surges on the highway at light throttle. Accelerates nicely in first and second gear, but is otherwise severely under powered. It basically won't accelerate in third fourth or fifth gear from a cruise. If you are shifting it at redline, third will be OK, but once you let off the throttle and try to cruise, it won't maintain speed. Here are some data captures that reveal what is wrong, but you have to analyze them carefully if you want to try to figure this out.

thecardoc3

To see the data captures better, right click on one and then "save target as". You can then drop them into any picture viewer and from there open them and zoom in. Be ready to take notes so that you can compare the data pids to each other.

stever

You forgot to link to the app that zeros in on the issue.

thecardoc3

stever said:

You forgot to link to the app that zeros in on the issue.

Seems there are no short cuts to knowledge, experience and the thorough application of common sense.

stever

Just need a coder.

If signal a is x, then y. Except if signal b is x, then z. Rinse and repeat until the discrete component or wiring or whatever that's out of spec is located.

You have all the screen shots from the various inputs. The next logical step is to have the program interpret the signals. If the tech insists on getting involved, I suppose the signals could be changed - i.e., what happens if the barometric pressure drops to 28.8? But why bother?

Mr_Shiftright

That might quickly get you to a certain broad condition, like "too lean fuel mixture" but I don't think it would take you very accurately to the "why"---the human brain still has to interpret.

stever

The point of the "next" step in the coding is to get beyond the simple gathering of sensor inputs that Doc's tool generated and implement the solution.

Here's a simpler example of how the reader could be coded to find the exact fault, and take the tech out of the loop.

And then let this tech do the heavy lifting.

roadburner

The problem is simple; the guy should have bought a Wrangler with the 4.0 I6- and dropped in a 4.7 stroker motor for around $3,000 if the stock 4.0 didn't have enough juice.
Turbocharging the I4 lump- what a pointless exercise in futility.

stever

That does bring up the issue of diagnosing modified equipment - do you bully ahead or first dial it back to "stock"?

thecardoc3

stever said:

That does bring up the issue of diagnosing modified equipment - do you bully ahead or first dial it back to "stock"?

Depends..

Mr_Shiftright

Any computer is only as good as the man operating it. I suppose every technician could have a Super-Cray in his toolbox, but....really?

In any event, you'd have to completely redesign the automobile to interface with a system like that, so welcome to Sci-Fi World.

I like those robots though. Maybe technicians will finally have clean break rooms!

thecardoc3

roadburner said:

Turbocharging the I4 lump- what a pointless exercise in futility.

But there is a silver lining. It made for a great case study in the basics of fuel control.

Mr_Shiftright

Well if the turbo kit was a fully engineering, dyno-tested, package that would integrate completely into the subject vehicle, that's one thing. But ordered parts from the Bolt-On catalog is yet another.

thecardoc3

stever said:

Just need a coder.

If signal a is x, then y. Except if signal b is x, then z. Rinse and repeat until the discrete component or wiring or whatever that's out of spec is located.

So what happens when an input falls outside of the parameters that the original software was written for, so it ends up buffered out of the scan data? Saying that another way, when the PCM on the car spits out data that is basically a lie, how is another software program going to be able to account for something like that unless the engineer writing it can anticipate every possible failure at every point in the system that could ever possibly occur?

stever said:

You have all the screen shots from the various inputs. The next logical step is to have the program interpret the signals. If the tech insists on getting involved, I suppose the signals could be changed - i.e., what happens if the barometric pressure drops to 28.8? But why bother?

That's the kind of flawed thinking that has helped create the shortage of qualified technicians, like the one's at the Honda dealer that can't figure out the heated steering wheel problem without help from the tech hotline. The sad part is how many are quick to blame the techs, instead of the system. Put the blame where it really belongs and train the techs correctly and fixing that steering wheel heater issue would have been a mundane event.

thecardoc3

Mr_Shiftright said:

I like those robots though. Maybe technicians will finally have clean break rooms!

Just think what it would cost to have some of those robots in a shop. The shop would have to be charging in the thousands of dollars per minute......

stever

I'm exactly blaming the system. Ideally the techs would be parts replacers.

Mr_Shiftright

True but the robot service writer will listen to your complaints very sympathetically for hours on end.

We all know where these robots are going ultimately---to war. We never got our atomic kitchens and flying cars and magic carpet pedestrian sidewalks either.

I won't say all futurists are wrong, but most are--except the very cautious ones who only go out 1-2 years, and they of course are already aware of the technology they are predicting.

stever

People want tech and that's the biggest thing they complain about when they don't get it or it doesn't work.

The tech techs should be the ones complaining loudest about their lousy tools.

Technology Problems Top a List of Car Complaints (NY Times)

Mr_Shiftright

A technician citing his tools as the reasons for his failure ? That sounds like a cop-out, doesn't it?

"If I only had a computer...." Leo Tolstoi

stever

I dunno - is it a cop-out that you have to have 20+ years of training and experience to figure out what those 16 screens Doc posted really mean? The tool should be able to tell you what they mean, not just spit back the results of the sensor readings.

Leo also said "We lost because we told ourselves we lost."

Mr_Shiftright

Well it would be a cop-out to claim you are a technician. I have no problem with dealerships being manned by "screen readers" but there had better be someone at hand who actually knows what they are doing.

stever

mmmm, guess I'm okay with having the balky "system" swapped out with another one. The robot can then box up the bad part and send it off to Central Processing for repair or recycling.

Mr_Shiftright

or put it in the refrigerator and then walk through a plate glass window.

thecardoc3

stever said:

I dunno - is it a cop-out that you have to have 20+ years of training and experience to figure out what those 16 screens Doc posted really mean? The tool should be able to tell you what they mean, not just spit back the results of the sensor readings.

That's all scan tools could ever do, display the data that the PCM sends to it. The scan tool can also "ask" the computer to turn on circuits on demand (bi-directional controls) for testing purposes.

BTW, this same exercise on a professional technicians site had immediate responses that nailed the issue and from there they have gone on to debate on what could be done to make this work. This is an example of the fundamentals in fuel control and at this level we are not talking twenty plus years experience, probably closer to five. There is much, much more that has to be learned during the journey to the twenty year mark and on. You should try dealing with systems that have over 100 data pids and seven or more computers that have to work together to make a system perform an intended task.

stever

Scan tools do more and more with each iteration. Like I say, you need better tools.

thecardoc3

I have more than $300K spent on tools as it is, scan tools, software licenses, are only serving to drive that total higher, with almost no hope of recovering the cost, never mind the thought of making money with them.

thecardoc3

stever said:

Like I say, you need better tools.

Well those captures started some nice dialog, but no one is trying to figure out how to interpret it. Techs do need better tools, but there is none better than the one in between his/her ears.

stever said:

I'm exactly blaming the system. Ideally the techs would be parts replacers.

http://forums.edmunds.com/discussion/39768/honda/pilot/half-completed-repairs-and-more-bad-news-2016-honda-pilot-long-term-road-test#latest

There has been pressure to make the job something other than what it really is for way too long. A lot of that has come from inside the trade as much as anywhere else. It gets marketed as a commodity and therefore people tend to treat it as such, but it is a craft and that Jeep exercise is an example of a minimal standard of training and knowledge that is required for someone to be good at it.

stever said:

You have all the screen shots from the various inputs. The next logical step is to have the program interpret the signals. If the tech insists on getting involved, I suppose the signals could be changed - i.e., what happens if the barometric pressure drops to 28.8? But why bother?

How about let's bother. What would happen if the baro pid was at 28.8"hg? What does the value in the data stream actually mean? How is that measured? How does that affect the other data pids shown, or does it?

thecardoc3

This describes the issue nicely.
https://www.youtube.com/watch?v=u6XAPnuFjJc

stever

The useless work part?

thecardoc3

I like the part about what happens to the reward/punishment aspect of managing the moment the job requires more than just manual labor. The job has always required more than just manual labor, even the simplest of tasks demand care and attention or else things can get ugly in a heartbeat.

Your statement "I'm exactly blaming the system. Ideally the techs would be parts replacers." is where the system is broken. The techs have to be anything but just parts replacers. That's why the flat rate system is failing the trade and consumer alike.