Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Autopilot saved our lives

This site may earn commission on affiliate links.
Status
Not open for further replies.
How is it that you don’t realize that ignorance of sleep disorders is not the root cause of the problem.

You seem to ignore the real cause appears to be your simple lack of good sense.

You said you were tired / ill because of various reasons and experienced several problems driving. Why didn’t you just stop and get some rest? You put yourself and others at risk in my opinion because you failed to act in a normal sane manner.

Take some responsibility for your actions. You’re old enough to remember when people actually considered that to be the proper course of action.
 
  • Like
Reactions: AdamMacDon
ignorance of sleep disorders
Ignorance is your problem.

Modern technology provides solutions for sleep disorders including better diagnostics, CPAP, oral fixtures, and medicines. Now we can add one more, AutoPilot. It adds an additional layer to vehicle safety. Perfect, no, but perfectible. We're alive and well and it was due to AutoPilot handling five unplanned micro-sleeps.

Bob Wilson
 
Ignorance is your problem.

Modern technology provides solutions for sleep disorders including better diagnostics, CPAP, oral fixtures, and medicines. Now we can add one more, AutoPilot. It adds an additional layer to vehicle safety. Perfect, no, but perfectible. We're alive and well and it was due to AutoPilot handling five unplanned micro-sleeps.

Bob Wilson

But I fear for a potential day where you might be dead and unwell where AutoPilot didn't handle a microsleep episode well. As another poster pointed out, what if your hand/arm gets heavy on the wheel and you kick it out of AutoPilot? That's only one of many failure modes both you and others experience every month with the system. Frequently.

I knew someone that suffered from a similar (same?) disorder. Except she didn't tell us before we took her out ATV'ing one day. She had an episode, torqued the steering over involuntarily, and flipped the quad. She ended up in the hospital but survived (and that's how we learned about her condition...).

I didn't care at the time she wrecked our quad. I wanted her to be ok. But damn, was I mad she risked her life and the lives of others knowingly. She understood this after, it was a bit of a wake-up call. Before that incident, she just accepted the risk of driving. Now, someone else drives.

I don't want you in her situation, but with a 4000lb chunk of steel moving at 100km/h down a road. It's not prudent. AutoPilot helps sometimes but is deeply flawed currently and absolutely cannot be depended on. Please, please find a less risky alternative.
 
As another poster pointed out, what if your hand/arm gets heavy on the wheel and you kick it out of AutoPilot?
For the nag, I use a sock with about 2 lbs of birdshot tied to the steering wheel rim. With practice: it is secure; located away from the airbag, and; full access to the controls. But my CPAP machine remains quite effective and I have a backup dental fixture that works although not as effectively.

One of the best 'home screening' techniques is a recording pulse oximeter. You wear it at night and the next day scan for low O{2} events. I used it to evaluate the effectiveness of my CPAP and the dental fixture.
I knew someone that suffered from a similar (same?) disorder. Except she didn't tell us . . .
My wife was narcoleptic, not so much since she has become homebound, arthritis. When I first met her, I had a Yamaha 550 with a sissy bar. She would 'go out' on the back and I would lean back to pin her against the sissy bar while slowing to park on the curb. After about 5 min, she'd come back and we'd complete our trip.

My wife's narcolepsy came on after puberty but was not diagnosed for 10 years. She suffered a lot from it before we met up, 42 years ago. I'd had a narcoleptic Aunt too. Regardless, sleeping disorders are common and treatable once they are diagnosed.

An excellent resource: Narcolepsy: Symptoms, Causes & Treatments – American Sleep Association

Narcolepsy research continues and we've visited: Sleep Division. This was a pivotal discovery in 2000:

The Stanford team, lead by Emmanuel Mignot, MD, PhD, associate professor of psychiatry and director of the Stanford Center for Narcolepsy, searched for defects in the brains and genes of narcolepsy patients from around the world. The team found that a small protein or peptide present in the brain cells of most people was absent in every brain of the people with narcolepsy that they studied.

"We think this is the cause of most human cases of narcolepsy that patients don't have this peptide in the brain," said Mignot. "Now we need to develop a drug that can go into the brain to replace it."

The peptide, called hypocretin, was first identified as a culprit for narcolepsy when Mignot and his colleagues found that dogs with narcolepsy had a problem in one of the hypocretin genes. Hypocretin was being produced in the animals but there was a breakdown in the communication system that allowed hypocretin-generated messages to be transmitted into the cell.

Surprisingly, the researchers found that people with narcolepsy generally have intact hypocretin genes. But they found that there is a malfunction further along the hypocretin production line. The peptide is not being produced where it is needed in the brain, where hypocretin seems to be responsible for promoting wakefulness

Bob Wilson
 
Last edited:
For the nag, I use a sock with about 2 lbs of birdshot tied to the steering wheel rim. With practice: it is secure; located away from the airbag, and; full access to the controls. But my CPAP machine remains quite effective and I have a backup dental fixture that works although not as effectively.

Statements like this is why most people are expressing a lot of frustration with you. It sounds bewilderingly untrue, yet others have attested to the fact you are indeed a real person. As an engineer you of course would know of all the cost vs. lives lost decisions that go into systems, and you're making a judgement that you are overall safer with Autopilot features. I understand where you're coming from here. However, that overall safety improvement is only true if you are also actively monitoring this system and not doing anything to make it inherently more dangerous.

AutoPilot, today, is absolutely a system which you must actively monitor. Tens of thousands of accident-free miles do not mean the system is safer than an unassisted human driver, nor does it account for everything an unassisted driver does. It's classic confirmation bias, which happens especially easy because the fact is highway driving is boring 99+% of the time, making a very high percentage of successfully handled miles. Even Tesla's own stats likely don't account for people likely disengaging AutoPilot before a dangerous situation, skewing their marketing numbers for miles safely driven. And I do really want to emphasise marketing numbers.

You've obviously put some thought into the sock thing to make it safer, but an emergency maneuver could very easily be resisted or altogether cancelled due to the jerk it will apply to the wheel when it moves. I don't say this so that you can find yet another way to defeat the "nag", I say it because the nag exists because you absolutely should be 100% attentive when current AutoPilot features are engaged.
 
. . . system which you must actively monitor.
Earlier this week, I realized driving on AutoPilot has become fun, almost a game. I score points by miles using just the turn signal and TACC nipple wheel. I get bonus points by recognizing an AutoPilot problem is coming up and transition to manual with no external effect.

Playing the AutoPilot game heightens one awareness and makes driving fun in a different way.

Bob Wilson
 
For the nag, I use a sock with about 2 lbs of birdshot tied to the steering wheel rim. With practice: it is secure; located away from the airbag, and; full access to the controls.

Please, please, please tell us you are joking. Right?

Multiple micro-sleep events, and you've disabled the "nag" that works to ensure you're still awake/alive?

It is too hard to believe that a sane, adult, human is writing these things. (And it may be too late for this year's Darwin Award submissions?)
 
  • Funny
Reactions: AlanSubie4Life
you are joking
Source: How much do you trust basic Autopilot?
  • 56.1% (46) - are comfortable with it
  • 29.3% (24) - use it but worry
  • 7.3% (6) - didn't purchase it
  • 4.9% (4) - would not use it with kids in car
  • 2.4% (2) - don't trust it at all
I am very comfortable with AutoPilot and more so after every firmware upgrade and my daily use.

Meanwhile, another source: Tesla Vehicle Safety Report
  • 3.27 million miles per accident with AutoPilot engaged
  • 2.19 million miles per accident with just active safety measures
  • 1.41 million miles per accident without any of the above
  • 0.49 million miles per accident per NHTSA
It is irresponsible to not use AutoPilot at every opportunity.

Multiple micro-sleep events . . .
Only if untreated or worse, undiagnosed in the ignorant. Source: Sleep Apnea | National Sleep Foundation

The treatment of choice for obstructive sleep apnea is continuous positive airway pressure device (CPAP). CPAP is a mask that fits over the nose and/or mouth, and gently blows air into the airway to help keep it open during sleep. This method of treatment is highly effective. Using the CPAP as recommended by your doctor is very important.

Other methods of treating sleep apnea include: dental appliances which reposition the lower jaw and tongue; upper airway surgery to remove tissue in the airway; nasal expiratory positive airway pressure where a disposable valve covers the nostrils; and treatment using hypoglossal nerve stimulation where a stimulator is implanted in the patient’s chest with leads connected to the hypoglossal nerve that controls tongue movement as well as to a breathing sensor. The sensor monitors breathing patterns during sleep and stimulates the hypoglossal nerve to move the tongue to maintain an open airway.

Bob Wilson
 
Meanwhile, another source: Tesla Vehicle Safety Report
  • 3.27 million miles per accident with AutoPilot engaged
  • 2.19 million miles per accident with just active safety measures
  • 1.41 million miles per accident without any of the above
  • 0.49 million miles per accident per NHTSA

This data does not tell us whether it is irresponsible not to use Autopilot!

As you know, here are some other questions this data does not answer:

1) Is using Autopilot safer than not using Autopilot, for a given drive?

2) Do Teslas have lower accident rates than other new cars? How about when using Autopilot as compared to other cars on the same roads?

3) Is Tesla safety improving with time? Has accident severity been reduced, if not the frequency (in the Tesla data set, the changes from quarter to quarter are very small)?

4) Are the Tesla “non-autopilot” miles gathered from a group of Tesla cars that match the autopilot miles, in terms of age/type/safety features?

5) What is the definition of an accident? Do fender benders count? Is it the same definition as NHTSA uses?

6) Related to point 5, is Tesla’s accident rate lower than the national average, when using the same definition? (Most likely yes, but how much lower?)
 
Source: How much do you trust basic Autopilot?
  • 56.1% (46) - are comfortable with it
  • 29.3% (24) - use it but worry
  • 7.3% (6) - didn't purchase it
  • 4.9% (4) - would not use it with kids in car
  • 2.4% (2) - don't trust it at all
I am very comfortable with AutoPilot and more so after every firmware upgrade and my daily use.

Meanwhile, another source: Tesla Vehicle Safety Report
  • 3.27 million miles per accident with AutoPilot engaged
  • 2.19 million miles per accident with just active safety measures
  • 1.41 million miles per accident without any of the above
  • 0.49 million miles per accident per NHTSA
It is irresponsible to not use AutoPilot at every opportunity.


Only if untreated or worse, undiagnosed in the ignorant. Source: Sleep Apnea | National Sleep Foundation

The treatment of choice for obstructive sleep apnea is continuous positive airway pressure device (CPAP). CPAP is a mask that fits over the nose and/or mouth, and gently blows air into the airway to help keep it open during sleep. This method of treatment is highly effective. Using the CPAP as recommended by your doctor is very important.

Other methods of treating sleep apnea include: dental appliances which reposition the lower jaw and tongue; upper airway surgery to remove tissue in the airway; nasal expiratory positive airway pressure where a disposable valve covers the nostrils; and treatment using hypoglossal nerve stimulation where a stimulator is implanted in the patient’s chest with leads connected to the hypoglossal nerve that controls tongue movement as well as to a breathing sensor. The sensor monitors breathing patterns during sleep and stimulates the hypoglossal nerve to move the tongue to maintain an open airway.

Bob Wilson

This dude is the KenM of the TMC forums!
There's no way he believes everything he's saying.
 
duty-calls.jpg


Regardless, it turns out my "nag defeat" device is 1 lb of birdshot in a sock:

sock_030.jpg

The scale already showed ~1/4 lb before measuring the shot bag.

sock_010.jpg

The typical wrap on the steering wheel. It can easily be taken off before giving others a Tesla test ride. Notice there is no interference with the air bag. Experimental results show only a slight right-hand bias with plenty of time to manually control.

sock_020.jpg

There is plenty of space between the sock and right-hand controls.

Bob Wilson
 
I've tested it too, Bob. And it doesn't always work. If you nod off at the wrong time you may not be as lucky as I was.

You replicated the problem with static object detection which sometimes works and other times, not. Which software version were you on?

So far, I'm finding on 2019.24.4, approaching stopped cars in traffic has worked every time and when following a car that stops. This has been consistent day or night.

My earlier testing has shown detecting the middle of semi-trailers often fails. Yet if you drive towards the rear wheels, it works and activates automatic braking. I'm waiting on 2019.28.x to do more regression testing:

Bob Wilson
 
Version prior to 24.4.
So far it has never missed detecting a stationary car ahead in my lane. I have seen a video of an M3 encountering barrels in an identical scenario as mine (although from the opposite side), and it merged across lane markers to get into the adjacent lane with no intervention by the driver. Seems like it sometimes works and other times not. Also, I have approached a car ahead of me stopped at a light and as I was slowing down (on AP) I got a collision alarm. There ahead of me about 75 feet or so, a couple inches outside the lane marker, were 2 plastic trash cans. The 2 cans and the car were the only objects around. I can only conclude that detecting stationary (non vehicle) objects only works part of the time.
 
  • Like
Reactions: bwilson4web
Status
Not open for further replies.