RomanTPA

Just in case anyone is interested

Ran 4 tests at speeds 15-25 mph. All have failed.

Failed 3ft by 8ft cardboard test


Failed 3ft by 8ft cardboard test


EDIT: So I did some research on PCS and grew some ***** and tried it out on a real person (myself):

Successful real human test

SNiiP3R

iTrader: (3)

It only works when you don't need it to work.

SRC350F

This ^^^ 100%

RomanTPA

Reminds me of this useless thing that can come up within first 5min of driving..

SRC350F

shoot.. they should provide us coffee :X

Dallasguy1

lol u should report this to Lexus USA and demand a free front bumper buffing haha.

GregCanada

I think it would be checking a heat signature or something similar, try a real person next time

lamikela1

It would be nice if we could see what the cardboard pedestrian looks like as I am not sure how lifelike this was. And no there are no IR sensors for this. It's all based on the front facing camera.

RomanTPA

The card board was standing straight up 8 ft tall and 3 ft wide.

dchar

I think the camera and radar are designed to detect vehicles and pedestrians only. So if the object, like your cardboard cutout, doesn't resemble either a person (shape) or a vehicle (metallic reflections) the system won't recognize it and will "fail."

xRYD3Rx

I think people shouldn't be driving a car if they need to have that feature in their car.
Seems useless to me probably $1k added to the msrp just for that option

RomanTPA

The way the system works is that it uses camera to recognize objects, such as vehicles and pedestrians. And it uses radar to compute density of the object. That's why even if cardboard looks like a pedestrian, it will not be recognized by radar. Check out this video:


It's part of F-Sport package. Soon every car will have this technology in it just like ABS and seatbelts..

spoogenet

Not exactly standard yet, but Toyota wants to make it such by 2018. I hear the point about driver capability, but think bigger picture for a moment -- a good computer system may not always be as accurate as a human, but it is much faster. Avoiding the collision is far preferable to minimizing damage when it occurs, but statistics dictates the best solution is to design for both scenarios.

The problem lies in false positives and how the system reacts. Back before independent testing, Volvo was a pioneer in production collision avoidance, but the IIHS doesn't rate their system so well. Volvo has tuned it a bit to reduce false positives, but the real question is -- did Volvo dampen it too much or did others design better for the test?

This tech is something you hope to never need and you hope it works when it is needed. Most of us will never test it intentionally.