I was scared to read about how a Tesla's door handles work -- it may apply to other modern cars, but I've only read about the Tesla. (It was probably when this incident first hit the news.)
The doors are electronic. The front doors have a mechanical override simply by pulling the door release lever further than usual. That's OK.
For the rear doors, there's an emergency cable to pull on at the very bottom of the door[1].
I was 11 when my dad crashed our car, at night at a decent speed. When we stopped moving, the car next to ours was already on fire, and I'm told ours also caught fire, although I didn't notice it at the time. My dad's (driver's) door was blocked, so he had to clamber across my younger sister in the front passenger seat. I unbuckled my younger brother (age 5), opened his door a bit, but had to kick it to get it to open sufficiently -- it was also badly dented in the collision.
If we'd had a Tesla, would you expect 11 year old me to know to "look below the seats for the carpet flap, find the little plastic tab, grab it and pull it to the side" -- and to be able to kick open the door while pulling that tab, since the side of the car is smashed in?
(We were all uninjured; some people in other cars had broken bones.)
Nobody would know. I am interested in Teslas, was driven in one (as a passenger only, sadly) and I didn't know.
A random person doesn't stand a chance.
Even if you know it as a trivia, it is likely that in an extremely stressful situation you will revert to doing what you have been doing all your life, but now with no effect.
A lot of airplane passanges die in crashes because they are unable to unbuckle the seat-belt (everyone is trained by cars to push on the button not pull the buckle).
Unbuckling airplane seatbelt is far easier then having secret door handle in the doors.
Phone lines are their own power grid. It can have outages too, but since they don't share infrastructure too much with HV lines, it's less likely both fail at the same time.
Nope, the ONT (Optical Network Terminal) that Verizon FIOS installed has a backup battery, but once that's exhausted it goes dark like the rest of the house. They also removed the copper line during installation, it seems like Verizon at least doesn't want to keep supporting their copper network where FIOS is deployed.
> Most of what was once a "trivial" design decision has become computer-controlled beyond recognition.
Indeed, it's almost caused by lazy thinking. If you need conditional behavior of a widget, the default is to reach for the almighty transistor, why spend more brain cells than that.
Yet... sometimes it's better to trigger your conditional behavior on some other physical phenominan (read: gas station nozzles[0])
More mechanical/moving parts increases risk and complexity, not reduces it. If something is misaligned or bends or breaks or even if it's just insufficiently lubricated, you're SOL. Solid state transistors are stupid reliable in comparison. Further, implementing more than two or three levels of conditional logic with mechanisms is prohibitively expensive, whereas you can have thousands of checks in a digital system practically for free. Finally, a mechanical system can only handle situations it is designed for, upgrades are difficult and sometimes impossible to implement. On the other hand digital logic is easily updated to deal with unforeseen issues.
Electricity is just as much a physical phenomenon as fluid flow or mechanical stress.
[Edit] I'm a mechanical engineer, no one should be any more enthusiastic about doing things mechanically than I am.
I guess the point is... say you use a microchip connected to a sensor to determine when to switch off a gas nozzle. It would work, of course, but if the sensor fails, or there's a bug in the firmware, or if power fails... gas spills everywhere. Whereas if you couple the trigger to a physical phenomenon that doesn't need electricity to work, now there's only one thing that can fail, and that's the part converting the physical phenomenon to a mechanical action.
I love computers as much as anyone on this site, but I like mechanical things too. I much prefer mechanical locks to digital ones. Mechanical locks are extremely reliable and well made ones work for centuries. Whereas I wouldn't trust a modern fingerprint padlock at all - what if the sensor fails? What if the batteries run out? What if moisture gets inside and corrodes the circuit board? One of the main jobs of a lock is to open when I need it to, and I just don't trust battery powered locks to do that.
That's the sticking point. People tend not to think of mechanical systems as systems full of many components any one of which can fail, but they are.
In the case of the gas nozzle, you have the orifice for the airflow which can get blocked, you have the diaphragm which can break, you have the connection to the lever which can sever, you have the lever itself which can bind. Critically if it does fail, it fails catastrophically - gas spills everywhere.
On the other hand consider a system employing a capacitive level sensor. When the sensor comes into contact with fluid, the dialectric constant changes and thus the voltage drop across the sensor increases to the point the valve will close itself. If anything goes wrong, say you lose power or a wire breaks or someone physically rips the nozzle off, the voltage still decreases below the cutoff threshold so the valve still closes - it fails safe.
People like mechanical systems because they are easy to visualize, but that's not the same as being simple or safe. At the same time, most of our experience with electronics is with low cost consumer electronics where the price of failure is low and there is strong financial incentive to use cheap components with short life expectancies in configurations that aren't fail-safe, so we tend to associate electronics with being unreliable.
But on a fundamental level, the force stopping two gears from passing through each other is exactly the same as the one producing a current when you apply a voltage to a piece of copper.
The problem is overengineering a device just for the convenience of the manufacturer or to add wow factor without taking in account reliability. It happens with mechanical and electrical devices as well. Tesla is not the only or the first to use electrically operated door releases.
It's unclear to me whether you are referring to the specific handle deployment or electronic components in general. If the former, I agree that it seems like a dumb feature but if people want to pay for it, so be it. However if you are talking about electronic systems in general, they are not overengineering for the convenience of the manufacturer nor wow factor. Components like solid state relays are chosen because they are safer and simpler than their mechanical counterparts - the fact that you can do more with them and they are cheaper is just further evidence of their overall superiority for most applications.
I think what a lot of people are forgetting here is that in the vast majority of situations you don't want random people to be able to easily open up your doors from outside. What failed here was not the door handle - which operated exactly as it should under normal circumstances, but the emergency override that should unlock the car when the airbags deploy. However, it's completely reasonable for some systems to fail when you crash at extremely high speeds - you can't make things indestructible. I see no evidence that a mechanical failsafe would have performed better.
Everything has its tradeoffs, the problem is what are you willing to give to get what, and I don’t think users know exactly that because systems have become so complex that even engineers sometimes struggle to make the right choices specially when pressured by marketing depts. To me personally i trust more a piece of steel wire than a motor-driver-microcontroller-switch-battery chain. Just for my door handles being able to move in and out.
Sure, because taking out the digital system to repair/upgrade it (unless your tap is somehow wireless too) is totally different from taking out these damaged/upgradeable mechanical mechanisms.
I think it's mostly due to the change in the makeup of the average "engineering" employee in corporations today. There's a very heavy emphasis on computer hardware and software knowledge, and much less on mechanical engineering and physical design.
If all you have is a hammer, everything looks like a nail. When you want to design a door handle with specific behavior and you understand software and electronics well but mechanical design less so, you design an electronically actuated door handle and do all the features in software.
It's cheaper, too, honestly, because revising a software system is inherently cheaper than the design-manufacture-construct-test loop for mechanical design.
But, as automotive design advances out of the technical design stasis it's been in for decades due to the dominance/barriers to entry of the US car industry, there's going to be a need for more advanced rules regarding what needs to be manual/mechanical vs. fly by wire.
A similar regulatory regime already exists in aircraft design, which could be a guiding example.
Modern day water systems will only work for a while, until the water towers run out of backup systems. Still need electricity to pump water up for gravity to pressurize/distribute it.
Not always. Also, a lot of water towers have alternative power sources they can use. More and more, solar panel arrays do this. They work very well despite the lack of constant power output because the water tower itself buffers the need for service.
In most places, the proximate cause of the pressure is gravity, which continues working in the face of a power outage. In many places, electric pumps have to raise the water to a tower or reservoir, but there's a buffer of water at height that can be drawn upon when the power is out. Yes, it'll run out eventually, but it meets the need in the short run.
In my experience, it's often true that city water remains pressurized even when a snow-covered tree branch has taken out a residential power line. Obviously, if the grid is taken out by a hurricane or earthquake or something the whole region is offline, but that's pretty rare here in the Midwest, but freakish ice/snow/wind storms knock down some wires and cause power outages at least once or twice a year. Why we can't convince regulators and the utility to spend 30% more initial capital and bury all the gross power lines underground where they're safe, instead of stringing them through the forest using dead tree carcasses instead is beyond me, but the city water treatment plant is typically powered by larger, more reliable segments of the grid.
However, now that I have a private well, yes, it kind of sucks when power goes out and water doesn't work. I've got 40 gallons of hot water in the heater that I can drain, and between 10 and 30 gallons in the pressure tank in the crawlspace. More than enough to draw some off for drinking and cooking, and a few buckets for pouring in the tanks of (gravity-operated!) toilets.
Speaking of cooking, another utility that's surprisingly operable in a power outage is natural gas. I can cook on my gas stove (but not in my electric oven), heat my house with my gas fireplace insert (but not my forced-air furnace), and heat water with my continuous-pilot water heater (but not fill it with my well pump) because the buried natural gas pipeline is always, always at 2 PSI.
> My modern sensibilities mean that I'm always surprised that toilets and taps work when there's no electricity.
That entirely depends on where you live! If you're in a valley and the water source is up in the mountain, gravity will do the job... but if you're living (in terms of height) above the water source, you'll likely run out of tap water very soon - depending if your water utility provider has installed backup diesel generators on-site.
And out in the country with my own well and pressure tank, I lose water as soon as the pressure tank is empty (about 10 gallons). Two or three flushes.
And very useful. We used them to power shortwave radios in '93, during the war in the Balkans to figure out what the heck was going on since there were communication blackouts.
Did you know in your boot (trunk?) if you have one there is a mechanical lever you can pull to escape that works at all times (car off, no battery, locked, etc.) Maybe by the bonnet (hood?) release catch there could be a similar 'unlock and open everything lever' for emergencies.
That trunk latch is there because of legislation introduced at the request of a couple who were kidnapped and locked in the trunk of a car. Before then, you couldn't open a trunk from the inside.
Nothing similar has happened with the hood or door handles...yet.
A mechanical release would be good, yes, and I agree that modern cars are too obsessed with drive-by-wire when it comes to safety features like doors and pedals. But that lever wouldn't have helped in this case.
I infer that the driver was unconscious and bystanders/first responders outside the car were unable to open the doors. There are no claims that the internal handles were defective, they only complain that the futuristic, flush external door handles that are intended to detect key presence plus an approaching hand and extend when you reach for them were inoperable. However, there are no details in the article on whether the doors were even unlocked or airbags deployed. Regulations require that the doors unlock after the airbag deploys, most manufacturers also turn off the ignition and fuel pump (in an internal-combustion vehicle); I expect that this routine would include extending the handles in a Tesla. However, Tesla handles getting stuck is a pretty common problem, it seems the standard response is to pry them open with a credit card or similar object. And typically it's the electric actuator and slide mechanism that jams, so other doors would have worked, in this case it sounds like the power to the door actuators was gone due to the crash so no door would open.
But car doors getting crushed in and becoming inoperable is a common problem, and the standard solution is a glass breaking tool or other heavy/sharp object. Had the door latches been crushed and stuck closed in the attorney's recommended "Chevy pickup", regardless of whether the handles were accessible, the police officer who responded first would have been expected to break a window. It sounds like the first responder got stuck in the problem-solving process when their plan to open the doors using the handles didn't work and they didn't know what to do. That's easy to identify from my armchair, but I'm sure it would be different when you've got an unconscious guy in a smoking Tesla in front of you, and certainly is no good resolution for the widow.
> Did you know in your boot (trunk?) if you have one there is a mechanical lever you can pull to escape that works at all times (car off, no battery, locked, etc.) Maybe by the bonnet (hood?) release catch there could be a similar 'unlock and open everything lever' for emergencies.
I would think that depends on the country. Every country has different car regulations, there is no universal car regulatory body.
I can't understand what you think this adds to the discussion? It should be obvious to anyone with common sense that there is no single regulatory body and regulations vary?
There is an old story in my family about my father buying the family car, the first to have electric windows/locks. In his mind those were a potential safety issue (winter, dead batteries). He insisted on at least the sunroof having a manual crank. That wouldn't help anyone trying to get in from outside, but at least someone on the inside could get out.
I agree, the handles on the Model S are a little too over engineered, but what prevented the first responder from breaking a window? How is this different than the doors just being locked? I own a Model 3, and the lever handles still wouldn't have opened when the car is locked.
I worry about this with the trend of using more laminated glass that is really difficult to break.
But this Model S is very unlikely to have had that, and it sounds like there was a policeman on the scene. Surely the officer would have access to something that he could have used to break the glass? I've read a few articles about this lawsuit and I have yet to see a clear answer as to why that wasn't the solution.
It makes me wonder if the "doors wouldn't open" part is just a spurious claim and the lawsuit is really about the post-crash fire.
According to the article it wasn't just random bystanders, it was a police officer.
"But Awan survived the crash, and he could have escaped the smoke and fire, too, Grossman said — if only the police officer who arrived on the scene could have opened the car’s doors."
Hey, one of my Scouts did this - rescued a family from a burning car. Because he had the wit to do it, while others stood on the sidelines. Probably because he'd been actually trained in taking quick action, in leading by example. He was 16.
So I can see that folks may be paralyzed by indecision, even by the intensity of their feelings. Ok, 'normal' people want to help but probably don't because of a lifetime of holding back.
I might, but I'd perhaps be more likely to risk my life saving someone from a Citroën 2CV or a VW Beetle. I'm only half joking. There is a real tendency to judge other people by what kind of car they're in and the decision whether to approach a burning car is likely to be made quickly rather than through careful application of logic.
If you regularly drive more than one kind of car you might notice that other road users treat you quite differently depending on which one you're driving.
In my car (not a Tesla) the handle opens the door through a mechanism that relies on a relatively thin steel cable. I've had this cable snap under normal circumstances (which is how I found out about this mechanism). I would not be surprised if this mechanism failed in a wide variety of circumstances. If my life depends on the door handle working properly, I'm a dead man.
Mechanical mechanisms are fragile - anything strong enough to survive impacts at highway speeds at an arbitrary angle is going to be too bulky to move manually without the assistance of yet another fragile mechanism. You can't design unsinkable ships, at some point you need life boats.
My goodness the entity which approved this death trap (if any) needs to review the requirement. The "ux designer" has failed miserably here. There should always be a fallback option. These issues with the handles have been reported in various tesla fan forums even. Very sad story.
Just seen the speeding mentioned, valid point , extremely irresponsible , however the topic is vehicle safety. When it comes to human life , every % to prevent death should count.
I'm already uneasy with the move from analogue interfaces to digital, but there's also this middle ground where "fanciness" just gets in the way of convenience.
Realistically, we're going to have a tough time finding out what dangers we'll have from moving from analogue dashboards to digital ones. Feels a bit like when we got rid of the incandescent bulb before LED bulbs were really ready.
Don't many modern cars automatically lock the doors once you exceed a certain speed anyway, so even if they could have pulled the handles the doors may have still been locked?
yes and no. Modern cars automatically (most with an option to disable) lock the doors, yes. This is to maintain crash integrity BTW. Not anti-carjacking as many folks seem to assume. But also, it's required that the doors mechanically unlock when you use the interior handle.
In this case:
> That’s because the car’s retractable door handles, which are supposed to “auto-present” when they detect a key fob nearby, malfunctioned and first responders weren’t able to open the doors and save Awan, the suit alleges.
It didn't matter. Clearly Awan was unconscious or unable to extricate himself. As you said, most cars auto-lock the doors. If the doors were locked (we don't know from this article), it wouldn't matter if the handles "auto presented" themselves or not. They wouldn't be able to open them from outside the car. If the doors were unlocked, well that's a different story.
This lawsuit is disgusting. The driver was doing 116mph in a 25mph zone. That was the cause of death. Nothing else. If you are doing nearly FIVE TIMES the speed limit, well, pardon me, but you're asking for death. It's a miracle he didn't kill anyone else.
He also hit 3 other cars, 2 trees, and rammed into a stone wall.
That this person committed crimes doesn't change that more reliable mechanisms should be preferred or required. That crimes were committed doesn't mean he should burn to death. The legal system should pick the punishment.
There's a reason you're not allowed to have pure drive-by-wire either, requiring mechanical connection to brake and steering.
While doors are generally less of a concern, but it's still nice to know that the door handle works when the electronics do not. Such as when there is a fire. Braking a window is harder than you might think if you don't carry a dedicated tool for it.
(And no, while this guy was a criminal, that does not mean all fires are by criminals, nor that criminals should burn to death)
All of that can be true, and irrelevant. If the car has a design that impedes rescue, it will impede the rescue of people who have done nothing wrong. Bad man is dead, buta lawsuit can serve as a nice smack to remind car builders about safe design.
I don't know anything besides what I just read in the article.
Sure, if you are going 5x speed limit and then crash, I wouldn't blame anyone but the driver for that, I could also understand that it could be reasonable to expect a fire in that situation.
But this article is about the doors, if you are unable to open the door without a working battery, that is indeed a problem...
FTA: Awan, a 48-year-old anesthesiologist and father of five, leased the Model S for two reasons, family attorney Stuart Grossman said: He was environmentally sensitive and safety conscious.
I'd say his family got a damn good lawyer, and by that I mean shameless.
Victim blaming here isn't really very helpful. Cars should be designed with an eye towards how they perform in an emergency. There is a reasonable case that Tesla didn't do that. As an owner of a Model 3 now, it is concerning to me how I would get out of the car if the electrics fail as exiting requires a button press controlled by a computer.
Most people don't have a hammer / seat belt cutter. It is not a required piece of equipment. People aren't trained in its use. There is no expectation that a standard driver (at least in the US) should be able to use this tool. This is Tesla's issue, not the drivers.
Fair point, there is an obvious assumption that the driver was capable of using such a tool at the time. Anyway, I intention wasn't to blame, merely an dark observation.
That's a pretty reductive assessment of what happened here. A system that requires you to remember to buy a window hammer on the off chance the doors don't open is a system that has failed. We expect cars to have thousands of other safety features — we don't bring our own seatbelts in case the ones in the cars fail, and we don't wear helmets in case the airbags fail to deploy.
I keep a small cutter + window breaker combo in my car. [0] not sure how effective it would be, but it's small and cheap. is it reasonable to expect people to do this? I dunno, you're driving around at high speeds in a metal box. no matter how well engineered, some fraction of the time it is going to crumple in a way that the doors no longer work. of course, if you're going to account for this tail risk, there's a lot of other emergency equipment you should probably carry around for more likely events. I can't leave my house everyday with 200 lbs of equipment.
Interesting, nearly all of my family and friends carry a window breaker in the car. I started doing so after trying to help someone out of a car after they crashed, and realizing just how tough those windows are to break.
No it's complete nonsense apart from public transport vehicles like buses where passengers are kind of expected to be able to find and use window hammers in case of emergency.
You must have never ridden public transportation. Must buses have certain windows with levers and need to be pushed out or kicked out. Trains have release levers to open doors and some have close buttons, at least the CTA does next to the red pull knob.
Having watched a fireman take three good swings with the pick side of a fire axe to break a tempered glass passenger window, I am thoroughly unconvinced that those hammers will do shit in an emergency.
That firefighter needed training. Trying to bash the window is not going to work. Wedging the axe between the window and door frame and using the axe as a lever with just a little force, the window would have easily shattered.
At least the hammers I see (on European trains and buses) have a sharp, conical tip. There are usually instructions to hit the window close to the corner.
I actually do agree that all drivers should have an emergency window breaker, irrespective of the handle fault issue. In the event of a crash where the car is submerged in water, the outside pressure might not allow you to open doors and window breaking or equalising pressure, is your only way out.
Having said that, GP made a rather disgusting comment where he is blaming a dead person for not having a non-mandatory emergency window breaker in a car compartment.
From the article description the handles don't appear to be fault-tolerant and should certainly be investigated.
supposed to according to whom? it's a small risk that you can partially mitigate by keeping a cheap tool on hand. it's up to you to decide whether it's worth the trouble.
> Awan, a 48-year-old anesthesiologist and father of five, leased the Model S for two reasons, family attorney Stuart Grossman said: He was environmentally sensitive and safety conscious.
It may be arbitrary but I did want to point out Teslas are not environmentally conscious, they're luxury vehicles [0] In order for electric vehicles to be anything but luxury their power sources will need to be environmentally sustainable and friendly as well.
Home automation, which I used to take gratuitous part of, is also a luxury. I figured out pretty quickly that for every physical interface I replaced, I still needed a manual override. For example: I could have eliminated wall switches when I was seeking to automate my lights based on presence and time. Instead, I chose some Zigbee wall switches that used an internal relay. Even if the house turned the lights off, you could still walk over to the wall and turn the light on. In infrastructure automation I generally have the same sort of concept just for different reasons. I imagine Tesla needs something similar.
Edit: Just did some further reading and manual overrides are present in Teslas. It's safe to say Dr Awan didn't know how to use it, or didn't think to use it in a panic. Education and accessibility sound key here.
The doors are electronic. The front doors have a mechanical override simply by pulling the door release lever further than usual. That's OK.
For the rear doors, there's an emergency cable to pull on at the very bottom of the door[1].
I was 11 when my dad crashed our car, at night at a decent speed. When we stopped moving, the car next to ours was already on fire, and I'm told ours also caught fire, although I didn't notice it at the time. My dad's (driver's) door was blocked, so he had to clamber across my younger sister in the front passenger seat. I unbuckled my younger brother (age 5), opened his door a bit, but had to kick it to get it to open sufficiently -- it was also badly dented in the collision.
If we'd had a Tesla, would you expect 11 year old me to know to "look below the seats for the carpet flap, find the little plastic tab, grab it and pull it to the side" -- and to be able to kick open the door while pulling that tab, since the side of the car is smashed in?
(We were all uninjured; some people in other cars had broken bones.)
[1] https://youtu.be/01lXcD_Uz74?t=41