Hacker News Reader

Comment by Abishek_Muthian | Hacker News Reader

I get why Tesla gets the flak from analysts, it being a public company; But I rarely see an analyst speak about Tesla's impressive record in building safe cars.

Especially when there are car manufacturers who try to hide flaws which can lead to catastrophic malfunction or had been caught with cheat devices for emissions test.

Also one more thing which is less talked about is Tesla's autopilot being useful for people with disabilities although Tesla hasn't added any specific features for people with disabilities AFAIK.

View on HN

But I rarely see an analyst speak about Tesla's impressive record in building safe cars.

Tesla's car safety features are not a major selling point compared to its EV features, and moreover are generally features common to EVs (i.e., additional crumple zones, floor rigidity).

Tesla's autopilot still struggles to identify white trucks against a blue sky, and highway dividers, and can't seem to tell when the driver is asleep and has his hands off the wheel. So crash safety features take a back page to known dangers of Tesla cars that make Teslas more likely to crash than other vehicles.

Especially when there are car manufacturers who try to hide flaws which can lead to catastrophic malfunction or had been caught with cheat devices for emissions test.

Literally every newspaper and news organization in the US and Europe covered the VW emissions scandal. People went to prison over it. Billions of dollars in fines were paid. The stock price was downgraded by analysts for months. Ford and GM's stock prices are also down significantly despite near-record profits due to missteps in the sedan market. Analysts aren't biased against Tesla, they're biased against any company mis-executing.

Don't forget Toyota and the saga of the mysterious acceleration.

IIRC, Toyota was eventually successfully sued because they did not follow industry best practices in developing the sw for the ECU. Any idea if Tesla's software is following, say, ISO26262?

Yes, that was a big issue too. Covered by every newspaper and media organization in the West, multiple investigations by safety agencies and private parties, and their stock was downgraded by analysts during this "saga."

Any idea if Tesla's software is following, say, ISO26262?

With respect to the console? Unknown. With respect to Autopilot? It's clear that they're not, since they're are too many issues and regressions that would have been caught if they were following those standards.

> known dangers of Tesla cars that make Teslas more likely to crash than other vehicles

Do you have any stats on this?

https://medium.com/@MidwesternHedgi/teslas-driver-fatality-r...

Whenever you see someone showing a brief and incomplete window of data, you have to take a step back and ask yourself "why?". There is seldomly a case for you to just disregard newer data when trying to predict into the future.

That being said, this is simply a Tesla short that used incomplete (but true in the past) statistics to try and make Tesla look bad. They time boxed their statistics to 2016 and despite the post being made in mid 2018.

Tesla for the years 2019, 2018, 2017 has had 7 deaths per year each (6 so far in 2019). Despite the number of Tesla's on the road showing hockey stick like growth, the number of fatalities has remained nearly constant.

If the blog post were to be true, the number of fatalities due to accidents in 2017 should have been 14, then 24 in 2018, and closer to 50 in 2019.

Just to approach this from another angle: Teslas currently make up about 0.1% (very very roughly) of all cars in the US. From that, you’d expect around 40 fatalities per year, and increasing fast.

I doubt the extrapolation is that simple. Most fatal wrecks happen to people who are unlikely to be owning an expensive EV.

Look at stats for other luxury vehicles. They have much lower crashes per millions of miles driven.

> make Teslas more likely to crash than other vehicles

Citation needed. Which other vehicles? Tesla's overall safety record is quite good. Every car has things it does well and things it does badly, and every other manufacturer gets judged on the safety record on balance.

I mean, do you regularly post on HN about, I dunno, Toyota's safety record given the high center of gravity of its SUV offerings that make them "more likely to roll over than other vehicles" and claim that "crash safety features take a back page" to that problem?

Last I checked, Toyota's cars don't roll themselves over. The driver of the SUV needs to be going too fast and take a turn too sharply to cause a rollover. Indeed, Tesla Model Ys, and all SUVs, are also prone to rollover due to having a higher center of mass than a sedan. This is why anti-rollover features are now standard in SUVs: to minimize the likelihood of a rollover occuring, and then to mitigate the harm to the occupants of the vehicle in the event of a rollover.

In contrast, Teslas literally drive themselves into trucks, stopped cars, highway dividers, etc., resulting in the deaths of 7 drivers so far in a base of only a few hundred thousand vehicles, versus a rest-of-the-industry statistic of 0 across hundreds of millions of cars.

And finally, Toyota doesn't go around bragging about how its vehicles are the safest cars ever, or misstating (or even outright lying) about IIHS or NHTSA safety tests.

If you mean 0 deaths per 0 miles vs 7 deaths per 1+ Billion miles that’s hardly impressive.

PS: Oddly enough of you replaced those billion miles with average American miles in average cars you would expect 12.5 deaths. Tesla’s autopilot might be less safe than other new cars in their segment especially if driven defensively etc, but that’s harder to quantify.

Well, teslas ARE more likely to crash than commercial airliners, because "everyone knows you are more likely to die driving to the airport than in the plane"

Other than that, tesla does publish some data:

https://www.tesla.com/VehicleSafetyReport

That said, people commonly constrain the Tesla comparisons to newer cars or in some other way.

See wikipedia's "aviation_safety" page. Avaiation is clearly safer per mile (a factor of 60 or so), but not that much safer per hour (factor of 4 per so), but is 3 times more dangerous per journey.

So driving to the airport might well be safer than flying across the country according to the journey metric. Even using the hour metric I often am less than an hour from the airport and take more than a 4 hour flight.

> Also one more thing which is less talked about is Tesla's autopilot being useful for people with disabilities although Tesla hasn't added any specific features for people with disabilities AFAIK.

Autopilot isn't solution for disabilities. If you aren't able to drive car without AP, you cannot drive one with AP - you have to be able to intervene and take over at any point of AP operation.

Don't get people killed by spreading misinformation.

I said autopilot is useful for people with disabilities and also added Tesla hasn't added any feature for disability assist. I didn't imply disabled should go buy a Tesla.

Diability can be of various nature, affecting accessbility in various manner. Iam myself disabled, Tesla is not available in my country; I hope to drive one when available and so I keep track of experience of disabled people with Tesla.

That said you can find numerous videos of disabled people testifying Tesla autopilot is easier to use[1].

[1]:https://youtu.be/cuZYROM4Vrs

AP is a convenience package - it makes car easier to use, in best case scenario. In worst case - it's harder - you need to take over and correct behavior to avoid crash.

I feel you, but don't put your life in hands of a convenience package.

Some disabilities are not black and white--you may be able to drive a car only 20 minutes at a time due to a disability, but with say adaptive cruise control/lane keeping you could drive it 60 minutes at a time.

That's pronbably because Tesla has an impressive record of overstating how safe their cars are. Notice how this is the first Tesla car to qualify for Top Safety Pick+? That's not because the Model S wasn't tested, it's because it didn't qualify - and Tesla went ahead and claimed it was the safest car on the market based on a dubious interpretation of the much more limited NHTSA tests which was rejected by the NHTSA themselves.

Related IIHS ratings where it failed to achieve the award criteria: https://www.iihs.org/ratings/vehicle/tesla/model-s-4-door-ha... https://www.iihs.org/ratings/vehicle/Tesla/model-s-4-door-ha...

Other cars like Chevrolet Volt and Toyota Prius scored better.

I agree with your thought process but you also have to realize that overtime issues can arise. It would be difficult to say 'look at how safe they are' when in months or years later issues can come at their doorstep. I think analysts don't want that on their record.

> car manufacturers who try to hide ... or had been caught

Tesla went for years trying to make the _sticker price_ include "four years of gas savings: $3,500", bringing the 3 to $31.5K.

Of course you couldn't buy one for that price. That was fairly deceptive. Then Tesla fans tried to spin it as "trying to bring honesty to the market by looking at TCO"... except while they were happy to mention savings from gas, they didn't mention the cost of installing 240V, and their electricity costs were somewhat... absent.

This always comes up, so I am curious if one fixes for "where the Tesla is driven" when doing safety analysis? Considering how difficult it is to get a Tesla (and for older models, how expensive), surely if one fixes for location, such as suburbs and well designed metropolitan areas as well as the type of people that drive Teslas, then perhaps other vehicles will do just as well?

I am by no means saying Tesla is less safe, in fact I do believe these claims, I am just not sure if they are truly thorough.

The second and third paragraphs of the above comment are bound to age poorly.

Reply on news.ycombinator.com