Team-BHP > Road Safety


Reply
  Search this Thread
41,862 views
Old 28th May 2013, 23:30   #46
BHPian
 
Tejas Ingle's Avatar
 
Join Date: Dec 2010
Location: Pune
Posts: 275
Thanked: 380 Times
Re: Airbags deployed after hitting a dog. Understanding why?

Yes Jeroen, you are right about tensioners. They are activated along with the seat-belts. But pre-tensioners are activated when the ABS or ESP kicks in. This happens in all the high end cars. I have felt the pre-tensioners first hand and I would say that the time between pre-tensioners and air-bags would be very distinguishable since a car may travel considerable distance after braking/swerving and before impact.
Tejas Ingle is offline  
Old 29th May 2013, 07:19   #47
Distinguished - BHPian
 
Join Date: Oct 2012
Location: Delhi
Posts: 8,036
Thanked: 49,760 Times
Re: Airbags deployed after hitting a dog. Understanding why?

I don't want to get caught in semantics, but are you saying there is a difference between tensioners and pre-tensioners? Either the belt is working normally, ie it locks/doesn't unwind or it gets (pre) tensioned.

For me, its one and the same:

A system that takes up the slack in the seatbelt in a fraction of a second. When it kicks in you will definitely notice so.

Most use some sort of 'explosive' gas, the older system are purely mechanical with in essence a spring.

The newest systems have some additional gadgetry that allows after tensioning a bit of slack again to reduce head/neck injuries further.

Actually, I think the term pre-tensioning is a bit misleading in the first place; There is actual tensioning happening when this system kicks in. But it is before your body starts actually moving forward. Maybe that's why the call it pre?


Jeroen
Jeroen is online now   (1) Thanks
Old 29th May 2013, 11:37   #48
Senior - BHPian
 
DerAlte's Avatar
 
Join Date: Mar 2007
Location: Bangalore
Posts: 8,069
Thanked: 2,919 Times
Re: Airbags deployed after hitting a dog. Understanding why?

Quote:
Originally Posted by Jeroen View Post
i could be wrong, but I don't think heavy braking will put a car in "crash mode". ...
Correct, deceleration is not fast enough to infer there is an impending crash.

Quote:
Originally Posted by Jeroen View Post
... Still, from all I read my two cents would be that the airbags should not have deployed in this particular case. So it was probably some sort of a technical glitch. The sensors or the computer or something. ...
Your (very appropriate) expression highlights the basic issues when comparing automation logic v/s human interpretation

1. Humans use accumulated information from past experiences, sensory data (from eyes, skin, ears etc.) and intelligence to analyze and infer in any situation

2. The current state of practically and economically feasible automation technology cannot replicate any of the above - neither the accumulated information, nor the sensory sophistication, and definitely not the intelligence

3. What seems a perfectly logical expectation to humans is actually a product of the above - the (assumed) simplicity of which belies the fact that there was complex processing in the brain to produce it. This complex processing cannot be (currently) replicated in any computing system cheap and robust enough to be fitted in a car

4. A human will see a dog or a handbag which triggers something (ignore the right or wrong of it) - is it possible for the Airbag sensor to compute the Size or Mass of the object hitting the car? That also in a timeframe that is suitable to 'act in time' (safety consideration)?

5. Any and every automation system works on the basis of mathematical equations, appropriately simplified for practical implementation, to achieve the control / activation objective. Should one build in complexity that might compromise the primary objective (in this case) firing the airbag to prevent / minimize injury? For what reason - save the cost of the replacement airbag? Isn't that cost much much less than the hospitalization cost due to injury?

Quote:
Originally Posted by Tejas Ingle View Post
Yes Jeroen, you are right about tensioners. They are activated along with the seat-belts. But pre-tensioners are activated when the ABS or ESP kicks in. ... I would say that the time between pre-tensioners and air-bags would be very distinguishable ...
Err... what is the connection between Seat Belt and Electronic Stability Program? Is the seat belt expected to hold you in the seat if you swerve the car??? -ive g produced due selective braking by ESP is way below the trigger threshold.

The time between seat belt (pre)tensioning and airbag firing can be detected only in an appropriately equipped test facility - the driver cannot figure it out a. because in that situation the focus is elsewhere and b. no one (other than an accomplished musician) will be able to detect a time difference of a couple of hundred milliseconds between 2 events separated in space.

Quote:
Originally Posted by Jeroen View Post
... there is a difference between tensioners and pre-tensioners? ... Actually, I think the term pre-tensioning is a bit misleading in the first place ... But it is before your body starts actually moving forward ...
Right on all counts, @Jeroen. 'Pre' is used by some to denote active belt tensioning by the 'tensioner' before the seat belt latch/lock engages.
DerAlte is offline  
Old 29th May 2013, 11:52   #49
BHPian
 
lucifer1881's Avatar
 
Join Date: Mar 2011
Location: Hyderabad
Posts: 368
Thanked: 2,030 Times
Re: Airbags deployed after hitting a dog. Understanding why?

On a related note, what would have happened if a car at 80kmph does not brake and hits a dog or a deer or some other animal? I am assuming due to impact the speed would decrease to say 50-60 kmph or even 20-30 kmph, and the car would keep moving. Would an airbag deploy in such a case? And would it not be extremely dangerous for an airbag to deploy when the car has not come to a complete stop? Do manufacturers build in any logic to not deploy an airbag if the car is moving above a threshold speed?
lucifer1881 is offline  
Old 29th May 2013, 12:01   #50
BHPian
 
Join Date: Oct 2012
Location: Bangalore
Posts: 262
Thanked: 114 Times
Re: Airbags deployed after hitting a dog. Understanding why?

Quote:
Originally Posted by lucifer1881 View Post
On a related note, what would have happened if a car at 80kmph does not brake and hits a dog or a deer or some other animal? I am assuming due to impact the speed would decrease to say 50-60 kmph or even 20-30 kmph, and the car would keep moving. Would an airbag deploy in such a case? And would it not be extremely dangerous for an airbag to deploy when the car has not come to a complete stop? Do manufacturers build in any logic to not deploy an airbag if the car is moving above a threshold speed?
Every manufacturer would have set a minimum threshold for the vehicle speed. For example: 30 kmph. Only if the vehicle speed is above this and the system detects a crash, the airbags would deploy. If the speed is less than the threshold, airbags won't deploy.

The exception is of course the pedestrian airbags which need to deploy even at lower speeds but that also is having its own threshold.
MotoNanu is offline  
Old 29th May 2013, 12:04   #51
BHPian
 
lucifer1881's Avatar
 
Join Date: Mar 2011
Location: Hyderabad
Posts: 368
Thanked: 2,030 Times
Re: Airbags deployed after hitting a dog. Understanding why?

Quote:
Originally Posted by MotoNanu View Post
...Only if the vehicle speed is above this and the system detects a crash, the airbags would deploy. If the speed is less than the threshold, airbags won't deploy...
That is not my question. I am asking would an airbag deploy if the vehicle has hit an object and is still doing, say, 50 kmph? In this case, the driver has absolutely no steering control and an airbag deployed in a moving car could potentially cause more harm.
lucifer1881 is offline  
Old 29th May 2013, 12:11   #52
Senior - BHPian
 
supertinu's Avatar
 
Join Date: Aug 2010
Location: Bangalore
Posts: 1,158
Thanked: 1,164 Times
Re: Airbags deployed after hitting a dog. Understanding why?

Quote:
Originally Posted by groom View Post
Not a coincidence, purely calibration grey zone!! ideally should not deploy.
The calibration should take care of "certain mass deflecting the sensor" kind of misuses.
Yea based on my experience i feel that the system should be smart enough to understand this. Maybe it exists and just that the car i was driving had a less advanced version of it.

Quote:
Originally Posted by PapaBravo View Post
When we are talking about deceleration, I could resist but crunch some numbers.
Thanks for sharing these numbers. This is exactly the math I was doing here and the reason why I felt that the system was over sensitive since the de-accelration caused by the dog is going to much smaller compared to hitting something anchored like a tree or a wall.

Quote:
Originally Posted by DerAlte View Post
@vikram_d, if I may comment on your close-to-perfect explanation:
I think you have explained the situation in my case very well and I can understand what exactly happened. The system in my case was not smart enough to assess the extent of the impact and since the system needs to respond before some real harm happens to occupants, it just went ahead and deployed airbags. However in reality the impact was very momentary.

Quote:
Originally Posted by Jeroen View Post
I don't want to get caught in semantics, but are you saying there is a difference between tensioners and pre-tensioners? Either the belt is working normally, ie it locks/doesn't unwind or it gets (pre) tensioned.
Jeroen
Its possible that I may be wrong in the symantics. All the seat belts lock when there is a rapid pull but thats a reversable process and they unlock upon retraction. however when I used term pre-tensioners it implies the active phenomenon that happens along with air bag deployment. Its a one time use thing just like air bags and after that you cannot reuse the seat belts. Atleast in my case in mazda3, after deployment I couldnt pull out the seat belts, they were stuck.

Quote:
Originally Posted by lucifer1881 View Post
That is not my question. I am asking would an airbag deploy if the vehicle has hit an object and is still doing, say, 50 kmph? In this case, the driver has absolutely no steering control and an airbag deployed in a moving car could potentially cause more harm.
Very good point. I highlighted this in my initial write up. In my case I hit the dog atleast at 60kmph or even more and once the air bags deployed my hands were off the steering. But I had kept my foot on the brake hence the car continued almost in a straight line and stopped. But had the steering swerved for some reason or something were to get stuck under the wheel, i wouldnt have any control.

Last edited by supertinu : 29th May 2013 at 12:18.
supertinu is offline  
Old 29th May 2013, 12:37   #53
Senior - BHPian
 
vikram_d's Avatar
 
Join Date: Jul 2006
Location: Bangalore
Posts: 2,427
Thanked: 1,185 Times
Re: Airbags deployed after hitting a dog. Understanding why?

Quote:
Originally Posted by lucifer1881 View Post
On a related note, what would have happened if a car at 80kmph does not brake and hits a dog or a deer or some other animal? I am assuming due to impact the speed would decrease to say 50-60 kmph or even 20-30 kmph, and the car would keep moving. Would an airbag deploy in such a case? And would it not be extremely dangerous for an airbag to deploy when the car has not come to a complete stop? Do manufacturers build in any logic to not deploy an airbag if the car is moving above a threshold speed?
Completely depends on what sort of logic is built into the controller. There will never be a proper or definitive answer for this question.

I had narrated my own experience of hitting a dog with my car and the airbags didn't deploy and to this date I don't know whether they should have or should have not. I didn't even have the time to get on the brakes.

Quote:
Originally Posted by supertinu View Post
However in reality the impact was very momentary.
Whether hitting a dog or a wall, all such impacts are momentary. They take fractions of a second and the ECU controller has to decide it's course of actions within those fractions.
vikram_d is offline  
Old 29th May 2013, 13:39   #54
BHPian
 
sharc_biker's Avatar
 
Join Date: Apr 2013
Location: Mysore
Posts: 127
Thanked: 338 Times
Re: Airbags deployed after hitting a dog. Understanding why?

If airbag deployment is based only on rapid deceleration exceeding a threshold ( assuming it to be the car's max stopping speed), why are impact sensors needed at all?

IMHO, impact sensors play a major role than deceleration itself.
Eg: if a car rams into a tree, there is rapid deceleration meaning there has to be considerable impact too. Hence the airbag deployment.

In case of Vikramd's case of hitting a dog, there just wasn't enough damage to trigger the impact sensors to deploy the airbags. Had the mass of the dog or the speed of the car had been big/high enough, it would have.

There are many cars in which upon frontal impact, the curtain airbags do not deploy unless the sensors located inside the door/frames sense side impact. Then how do these work if airbag deployment was dependant on rate of deceleration?
sharc_biker is offline  
Old 29th May 2013, 13:42   #55
Senior - BHPian
 
supertinu's Avatar
 
Join Date: Aug 2010
Location: Bangalore
Posts: 1,158
Thanked: 1,164 Times
Re: Airbags deployed after hitting a dog. Understanding why?

Quote:
Originally Posted by vikram_d View Post

Whether hitting a dog or a wall, all such impacts are momentary. They take fractions of a second and the ECU controller has to decide it's course of actions within those fractions.
True, I meant more in terms of the amount of energy it pushes into the car. IN case of dog its just a fraction vs against a wall the entire kinetic energy of the car cause of the momentum is getting transferred in. But as concluded since the trigger event is in the beginning the system need to have sufficient smartness to be able to distinguish some how.
supertinu is offline  
Old 29th May 2013, 13:49   #56
Senior - BHPian
 
vikram_d's Avatar
 
Join Date: Jul 2006
Location: Bangalore
Posts: 2,427
Thanked: 1,185 Times
Re: Airbags deployed after hitting a dog. Understanding why?

Quote:
Originally Posted by sharc_biker View Post
If airbag deployment is based only on rapid deceleration exceeding a threshold ( assuming it to be the car's max stopping speed), why are impact sensors needed at all?
How did you arrive at this conclusion? I don't think anybody on the thread has said that rate of deceleration is the only criteria. There are a number of criteria which have to be met for airbag deployment. Side airbags have different sensors on the doors which play a role in deciding whether the side airbags should deploy or not.

All I will say is that airbags are complex systems which rely on a number of parameters to decide whether to deploy an airbag or not, but them being complex does not mean that they are foolproof or that they will work the way we expect them to. They will only work in the way that has been programmed into their controller.

Quote:
Originally Posted by supertinu View Post
True, I meant more in terms of the amount of energy it pushes into the car. IN case of dog its just a fraction vs against a wall the entire kinetic energy of the car cause of the momentum is getting transferred in. But as concluded since the trigger event is in the beginning the system need to have sufficient smartness to be able to distinguish some how.
Ah ok. I took the literal meaning of the word momentary. My bad.

As DerAlte has said above it is extremely difficult to build human intelligence in to a computer system. So airbag/car manufacturers err on the side of caution. Better be safe than sorry.

Last edited by vikram_d : 29th May 2013 at 13:54.
vikram_d is offline   (1) Thanks
Old 29th May 2013, 14:33   #57
Senior - BHPian
 
DerAlte's Avatar
 
Join Date: Mar 2007
Location: Bangalore
Posts: 8,069
Thanked: 2,919 Times
Re: Airbags deployed after hitting a dog. Understanding why?

Quote:
Originally Posted by MotoNanu View Post
Every manufacturer would have set a minimum threshold for the vehicle speed. ...
No relationship with speed. The video that @vikram_d posted shows that, doesn't it? It wasn't a simulation or Photoshop modification - the car was stationary when the airbag deployed.

And you meant 'passenger airbag', not 'pedestrian airbag', no?

Quote:
Originally Posted by lucifer1881 View Post
... In this case, the driver has absolutely no steering control and an airbag deployed in a moving car could potentially cause more harm.
Err ... why would the driver lose steering control? Forget to steer in panic is different! Also, the harm due to airbag deployment is - at any speed - much less than the harm it's absence would have caused, right?

Quote:
Originally Posted by supertinu View Post
... after deployment I couldnt pull out the seat belts, they were stuck. ...
In that state of mind (understandable; you cannot say you were cool as a cucumber throughout the episode, can you?), you might have pulled it quite hard. Happens to anyone in a hurry - that is the nature of the friction-based lock. The correct action is to unlatch.

Quote:
Originally Posted by vikram_d View Post
... I had narrated my own experience of hitting a dog with my car and the airbags didn't deploy ...
A lot depends on the point of impact and the size of the object. A smallish dog taking evasive action at the last moment may roll under the chassis or get hit like (TT/tennis) topspin stroke. In @supertinu's case the dog was large enough that it's torso was at the same level as the number-plate screws - and the impact was substantial. Not enough to deform the bumper, but enough to dislodge the number plate and for the sensor to act.

Quote:
Originally Posted by sharc_biker View Post
If airbag deployment is based only on rapid deceleration exceeding a threshold ( assuming it to be the car's max stopping speed), ...
Deceleration = (-ive) rate of change of speed, not speed itself.
Quote:
Originally Posted by sharc_biker View Post
... why are impact sensors needed at all? ...
That sensor is not sensing 'impact' in the sense you are using it when you say "Eg: if a car rams into a tree". A sensed deceleration of say 10m/s2 is the same whether the chassis went in 2mm or 200mm and/or whether the car travelled 2m or 200m. And the deceleration is not constant - it changes over the duration of the incident: a profile can be sensed. Please refer to the graph appearing earlier in the thread.

Quote:
Originally Posted by sharc_biker View Post
... upon frontal impact, the curtain airbags do not deploy unless ... airbag deployment was dependant on rate of deceleration?
Not difficult to understand. The individual sensors measure deceleration associated with their specific axis (along or across the car). On diagonal impact, if the deceleration profile was indicative of danger on both axes, *both* will fire. That is why airbags will not fire if the bonnet is slammed shut - that axis is not sensed.

Quote:
Originally Posted by supertinu View Post
... system need to have sufficient smartness to be able to distinguish some how.
Boss, the Airbag system of such smartness will cost as much as the car. Or, the car will cost upwards of 60L OTR in India (Volvo, for example)!!! If you can understand why passenger switch is not provided in low-cost cars, you should understand the impact of cost v/s price of car.
DerAlte is offline   (2) Thanks
Old 29th May 2013, 14:46   #58
Senior - BHPian
 
vikram_d's Avatar
 
Join Date: Jul 2006
Location: Bangalore
Posts: 2,427
Thanked: 1,185 Times
Re: Airbags deployed after hitting a dog. Understanding why?

Quote:
Originally Posted by DerAlte View Post
A lot depends on the point of impact and the size of the object. A smallish dog taking evasive action at the last moment may roll under the chassis or get hit like (TT/tennis) topspin stroke. In @supertinu's case the dog was large enough that it's torso was at the same level as the number-plate screws - and the impact was substantial. Not enough to deform the bumper, but enough to dislodge the number plate and for the sensor to act.
Very true and I am glad that the airbags did not deploy after I checked out the damage to the car. My only point in narrating that incident was to show that airbags will not always deploy or not deploy the way we want them to.

In the case of my hit the dog was the size of the full grown strays that we regularly see on our streets but it was pretty bulky in girth and the impact was dead center on the bumper. Thankfully the dog survived the impact without any harm.
vikram_d is offline   (1) Thanks
Old 29th May 2013, 15:00   #59
BHPian
 
Join Date: Oct 2012
Location: Bangalore
Posts: 262
Thanked: 114 Times
Re: Airbags deployed after hitting a dog. Understanding why?

Quote:
Originally Posted by DerAlte View Post
No relationship with speed. The video that @vikram_d posted shows that, doesn't it? It wasn't a simulation or Photoshop modification - the car was stationary when the airbag deployed.

And you meant 'passenger airbag', not 'pedestrian airbag', no?
I don't know about the general requirement, but few manufacturers do have the speed requirement. Only after the minimum threshold, the airbag deployment is expected.

Here is one of the link stating the same:
http://www.toyota-global.com/innovat...ve/airbag.html

As I said, this limit/ threshold differs from OEM to OEM but as I have seen many of them do specify the minimum speed.

I quoted pedestrian airbag as they are required to deploy at lower speeds and their threshold is different than the one for the driver/ passenger airbags.
MotoNanu is offline  
Old 29th May 2013, 15:08   #60
BHPian
 
lucifer1881's Avatar
 
Join Date: Mar 2011
Location: Hyderabad
Posts: 368
Thanked: 2,030 Times
Re: Airbags deployed after hitting a dog. Understanding why?

Quote:
Originally Posted by DerAlte View Post
...Err ... why would the driver lose steering control? Forget to steer in panic is different! Also, the harm due to airbag deployment is - at any speed - much less than the harm it's absence would have caused, right?...
Driver would lose steering control because his/her hands would automatically be jerked off the steering on deployment.

And, no airbag deployment at any speed is not always less harmful than its absence. If an airbag deploys at, say, 60 kmph on hitting a dog and the car is on a curve then it will go off the road. If this happens while driving in the mountains, then the car can even go off the cliff.
lucifer1881 is offline  
Reply

Most Viewed
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Copyright ©2000 - 2024, Team-BHP.com
Proudly powered by E2E Networks