16 ms·
I've worked for a large OEM, dealing with a large Japanese megacorp that is not Mazda for about two years (actually Mazda was one of our customers too, but I di
by mirkules 3y ago
I've worked for a large OEM, dealing with a large Japanese megacorp that is not Mazda for about two years (actually Mazda was one of our customers too, but I didn't get to work with them directly). This does not amaze me anymore.
We spent months agonizing over an interior temperature sensor, which was only used to display the information to the user on a smartphone app. We built both the hardware and software, and it was offered as an add-on at the dealerships. After months of negotiations, after the hardware was already built and the packages assembles, they decided temperature sensors were too inaccurate (+/- 5 degrees F) to use, and that it could present a legal liability. Again, this was nothing else but displaying the information on the app - and the user could then make a decision whether to remote start the car to cool it or heat it (no automatic process took place either).
This was at the height of "unintended accelerator" issue in Toyotas, so everyone was walking on egg shells playing it ultra safe to not invite any more lawsuits.
What surprises me is that this culture of "playing it safe" remained to this day, some 10 years later (but maybe it shouldn't).
- hirundo 3y agoMy cheap home thermostat has that frustrating +/- 5 degrees F accuracy. Is it very difficult to build an inexpensive 1 degree sensor?
- harrisonjackson 3y agoI am curious what an "expensive" one would actually cost, too... It is a car so already a large purchase. I'd pay a bit more for an accurate thermostat.
- mirkules 3y agoIt's not just material cost, probably different interface to the sensor so factor in some R&D, approvals, etc. Any time anything, no matter how small or innocuous (bracket, cable, screw, some piece of plastic cover, etc), was changed on a vehicle, it meant a different part number, which meant 6-12 months delay. This is because it has to go through all the testing - usability, fatigue, safety, etc all over again. This is why they pick cheap parts - not because they're cheap, but because they are old and got cheaper over the years, because old = already approved, which makes the lead time a lot less.
- vGPU 3y agoApparently +/- 2 degrees is fairly common. One of the problems is the heat from the device itself, as well as limited airflow creating localized hotspots.
- jacquesm 3y agoIt is if you don't want a calibration step. If you do calibrate then it's no longer an inexpensive part...
- radlad 3y agoI wonder if they were concerned someone would use the data to make a decision about leaving a living creature in the vehicle.
- PvtGleb 3y agoor medications or food. I can totally understand the legal liability angle 5° F is not a small margin of error.
- mirkules 3y agoNow that I think back, I believe you are correct. The decision was based on making a previously-unavailable feature available, which meant people could potentially (be irresponsible and) leave pets/children in the vehicle while relying on an inaccurate sensor, which could open them up to lawsuits.
- mardifoufs 3y agoWoah I'd have guessed that temperature sensors would be more accurate than that! Is it just an issue of cost, or are most affordable temp sensors that inaccurate and I've never realised it? That would explain a lot though!
- AlotOfReading 3y agoNo, it's pretty easy to get better sensors than that. E.g. a cheap-ass SHT20 will do +-0.3C. In fact, my own automotive parts recommendations are the next grades up (SHT21/SHT3x) as standard for my employer's boards because the cost difference is justified. Never underestimate the ability of a manufacturer to select subpar parts to save 25 cents on the BOM and spend 6 figures elsewhere trying to fix the resulting issues though.
- throwaway837367 3y agoNo, you read the datasheet wrong. It's +/- 3C.
- AlotOfReading 3y agoI'm looking at the datasheet right now. Relative humidity accuracy tolerance is +-3.0, temperature is +-0.3. Different tables.
- chris_va 3y agoI imagine the temperature gradient across the car might be +/- 3degC, ignoring the actual sensor.
- HankB99 3y agoMeasuring temperature is not trivial. There's convective, conductive and radiative heat transfer and they all factor into the measurement. And that may not accurately reflect the "cabin temperature," particularly in a parked car.
- jdjdjdhhd 3y agoIt should remain, at least for mega corps
- Aeolun 3y agoLiterally all of corporate Japan is built on playing it safe. Any innovation that comes out of it seems more like a happy little accident.
- Affric 3y ago5 Rankine is 2.55555555... Kelvin for anyone wondering which is there or thereabouts in the range of an average air conditioner.
- rainbowzootsuit 3y agoHenceforth I will be making my fill of liquid helium in my car. /s
- Affric 3y agolol… it’s also a hint that google will evaluate the reported conversion as one of pure scaling without any of the transformation that occurs in converting 5 Fahrenheit to 5 Celsius.
- wutwutwat 3y agoIdk about everyone else but when it comes to anything running in my car, _anything_, there is no such thing as excessive "playing it safe". It's a 2 ton mass of steel barreling down the highway at 70+ mph next to other unpredictable 2 ton masses, please for the love of God, fight to maintain that culture of "playing it safe", regardless of what you're working on and for what purpose.
- ipaddr 3y agoWould you be willing to pay millions for your car to make it safer? The pope has bullet proof glass, different body materials can protect your life. How would you define excessive?
- gottorf 3y agoYeah, everything is a tradeoff that should have a coherent risk-benefit analysis attached to it. "X at all costs" isn't realistic.
- wutwutwat 3y agoNever said "X at all costs" but thanks for trying to speak for me. Going forward, please note that my preference is to speak for myself, as should you.
- gottorf 3y agoI'm not sure why you took such offense; it's a reasonable interpretation of the words you spoke for yourself: > there is no such thing as excessive "playing it safe" As you noted in your other comment: > I'm asking them to "play it safe" when developing components for the car so the car doesn't kill me while I'm in the car. As in the old adage in computing ("the only unhackable computer is one that isn't connected to anything"), there's no way to ensure that the components of a car don't fail, even while in routine use. There is only more or less likely that they won't fail, and of course, less and less likely to fail is more and more expensive. We might say that the only uncrashable car is one that sits in the garage and never goes anywhere. Obviously, that would be playing it safe excessively, since it would defeat the purpose of having a car to begin with. But what about less obvious cases? Toyota recalled millions of cars for their "unintended acceleration" issue. The merits of that particular case aside, how much more would someone pay for a Corolla that would be progressively less likely to have safety issues? At some point before infinity, it would be considered excessive. I think the sliding scale of how safe is playing it too safe is a discussion very much worth having.
- glimshe 3y agoI'll be honest here... +/- 5F isn't useful to me as an add-on temperature sensor, that would have pissed me off as a consumer.
- kayfox 3y agoI work at a network hardware vendor who has a large Japanese customer base. After Fukushima we were asked to provide specifications for acceptable operating environment radiation levels, after some negotiation they relented and gave us a figure we could test at, so engineers drove hardware to a spot with high background radiation and ran it for a couple of days off a generator to test. The Japanese customers also insisted that each display on the hardware be the same shade of white, so they would all look nice in a datacenter, so our LCD displays have specifically graded white LEDs.