Safety Recall J0V - Where is the Remedy?

In the past, Toyota has been fined for not notifying owners about issues they could not fix,
So now they do notify us immediately. And start looking for a solution.

 

oh, they have a solution, but they don't want to spend the money.
looking for a cheaper solution...

 

Usually when you owe taxes, you wait for the last day to write a check to the IRS to pay it off. Similar for recalls like this.

 

You've looked at what this one is about, right? An edge case in the software that was supposed to be handled by going to failsafe/limp-home mode, but can end up going to immediate OFF instead.

Doesn't seem like a huge maze of possible solutions to choose from; they need a revision to the software that doesn't do the wrong thing there. And they have to test that, thoroughly, and convince themselves that change isn't going to introduce a new bug somewhere else. Those are the things that make all software maintenance hard, and everything that's hard about software maintenance is really hard when it's real-time, embedded systems control software.

And deep in your heart of hearts, you probably really do want them to spend enough time testing it before it is flashed into your car.

-Chap

 

not me, you should see the pile of dusty recalls on my desk

 

I agree. As an industrial electrician, if I made changes to a machine's programming, the programming software let me run it in test mode for as long as I wanted. Once I told it to save the changes, it warned me that mistakes could cause damage, injury or death and asked if I was sure. If I said "yes," it asked if I was very sure. Then it asked one more time if I was absolutely sure. Programming changes often have unintended consequences as those of you who have been through the recent Windows upgrade snafus have experienced. But snafus with car programing have much more severe consequences than on your laptop.

 

I spent most of 1983 to 2002 as a maintenance programmer, taking code that already existed and making it more correct/more full featured/used for a new purpose. (Think ghostwriter, except for programming languages) I would often contact my vendors with the following conversation "Hi I am getting this Bug, can you duplicate it?" "You can? Here is the fix for that." Then they had to prove that I had not introduced more errors than I fixed. It was usually a full year after I 'gifted' them with code before they published my work. Eventually I was in every years updates, I was making more changes than their in house programmers. (I had a real business to run, so my changes might only apply to my firm, or might be generally useful) By giving my modifications back the the vendor (for a nominal price) I did not have to re-modify every release they made.

If I was Toyota, I would QA the heck out of any changed code. I bet they QA until the Feds require them to 'solve' the recall problem. (I have never programmed for Toyota, and never written automotive real time code. I once wrote COBOL and BASIC for Honda Canada so Dealers could communicate with the Distributor, before the internet. Parts Ordering, Warranty Claims, New Old Stock Sales, Equipment Transfer, Transportation Inquiry) On Tuesdays, we ate curry goat rotis for lunch with the Toyota Programmers at The Roti Hut.

The Roti Hut | The Best Roti in the GTA

 

I’m not going to jump on the band wagon when they (Toyota) implement a supposed software cure for this converter anomaly. My plan is to hold off until the feedback starts rolling in about the adverse effects of this Bandaid.

 

I almost want to go take them up on the rental car. But I worry about what condition my hybrid battery would be in if it took more then 90 days for them to come up with a remedy.

 

Some things are easily solved in software, some less so. When the issue to begin with is a software unhandled failure case where the car turns OFF instead of going to limp-home mode, software seems like probably the reasonable place to correct the oversight.

Now, as a side issue, as covered here, they did discover that maybe 36,000 cars that supposedly had the earlier, E0E or F0R recalls done, actually didn't have that work completed at that time. So those cars (about one-in-twenty-ish of the cars subject to this recall) are still at greater risk of a failure, and the dealers will be expected to correctly re-do E0E or F0R on those cars when they come in.

So those one-in-twenty or so might end up noticing peak-power issues, if they're really that noticeable, after going in for this recall, and probably blame this recall for it, people being how they are. (I bought my Gen 3 with the earlier recall already done ... correctly ... so I've never had a chance to notice if it impaired power compared to before. Seems to drive like a Prius as far as I can tell.)

If you have a car where supposedly the earlier recall was already done, you can check your firmware versions to find out if it's one of the 36,000 that might not have been done completely.

-Chap

 

Mine had both the reflash and an inverter replacement done before I got it. Reflash done at 168k and inverter at 174k. I bought it at 209, and it now has 217k. The biggest fear I have is that any repair/reflash could have an obvious change/negative affect on the fuel economy or performance that my car does at this time.

 

I keep seeing that expression, and am somewhat fascinated by it. I'd like to think they're not always negative, too.

 

I'm trying to remember if I've ever had an unintended consequence in some code I wrote that I ended up liking better than what I had intended for it to do. In my corner of the field, that's kind of rare; the odds are on negative.

Now, people who get to play in AI / machine learning probably do get to watch their software succeed in ways they didn't expect. I bet that's fun.

-Chap

 

I agree it is rare, I can only think of a couple times the specs I was writing to were 'better' than the customer intended.

I wrote communications code for Honda Canada in 1987 - 1988. One feature the distributor wanted was syntactical error checking on the Warranty Claims form. The distributor was having to process dealer claims repeatedly until the dealership employee filled out all the fields in the paper form to suit the distributor. Now my software made the mechanic at the dealership answer all the questions to suit the warranty claims department at the distributor before he/she could leave the page. That was the intended result, and I accomplished it

The unintended consequence? Dealers got paid for warranty claims an average of 3 weeks quicker. Without all the re-work and mailing of the paper forms, the process completed in one day much more often. My software was a financial boon to both groups, even though the goal was just to save labor costs at the distributor.

Another 'feature' I used was debug announcements each time I went into a subroutine, read/wrote to disk, or read/wrote to the network. I was using that to find bugs during my development cycle. You normally turn off the various levels of logging once the program is done in development. I accidentally showed Honda Canada that I could monitor the network activity for the 16 dealer modems, and they insisted I leave network I/O debugging enabled in production. I wrote a viewer so the computer operators could see where in the phone call each dealer had made it to. At some level I was shamed that I never got to turn off my debug code because the Distributor wanted to spy on their Dealers.

Jimbo, your code NEVER made it out of debugging. (And the customer loved it)

 

I think that's next to unheard of in programming. In life decisions, it does happen sometimes.

 

Thanks to a heads-up from @Hopopotamus, it appears that the remedy is now available.

The J0V remedy notice doesn't seem to have shown up on NHTSA's site yet, but can be found at Toyota, Lexus & Scion Recall Lookup (I've attached a copy here).

The technical instructions are here.

For the technical instructions for the Prius v, see the post in the v forum.

Edit: looking over the technical instructions, one could get the impression Toyota was really embarrassed about the E0E/F0R recalls not always completing all the intended reflashes (because of the step missing from those technician instructions). Maybe some of the extra time coming up with this remedy really went into their latest, greatest Techstream version and its new, CUWC "update wizard" plugin, that now automatically runs multiple times as needed to reflash all the affected ECUs, and then ends with a step where Techstream itself verifies the installed calibration IDs in the car and reports them to the national database.

Seems like they really, really didn't want to hear about any more partial updates....

-Chap


about the toyota web sites: the toyota.com/owners site works pretty well if you allow javascript to run from toyota.com, jquery.com, and driverslogin.com. Toyota's other site (for recall lookups by VIN or state and license number) is some truly gold-plated javascript hellhole that still doesn't work after enabling seventeen different-origin scripts (which is where I gave up, leaving fifty-six of them still disabled, many of which I would not enable if you paid me).

 

Has anyone had this done? I have no idea what software version my 2012 is on and figured this may be a good thing to do. Wanted to know if anyone had done it and had any negatives or positives to report. Once it stops snowing and freezing below 0 here in the AZ mountains I'll schedule an appointment with the local dealer.

 

Have you got any letter yet. I've got one, saying they're working on a fix, and that there'll be a second letter when they're ready with the fix.