Current Accident Analysis and AEB Evaluation … 03... · Current Accident Analysis and AEB Evaluation Method for Pedestrians in Japan July1st, ... (d) Light passenger car 26 % (c)

Current Accident Analysis and

AEB Evaluation Method for

Pedestrians in JapanJuly 1st, 2014

National Traffic Safety and Environment Laboratory

Kenichi Ando

Final AsPeCSS Workshop

Outline

� Pedestrian accident in Japan

� Regulation and JNCAP (passive)

� Pedestrian accident analysis

� Evaluation method of AEB

2

Traffic Accident in Japan

0

200.000

400.000

600.000

800.000

1.000.000

1.200.000

1.400.000

0

2.000

4.000

6.000

8.000

10.000

12.000

14.000

16.000

18.000

1990 1995 2000 2005 2010 2015 2020

Acc

ide

nts

an

d In

jure

dA

ccid

en

ts a

nd

In

jure

dA

ccid

en

ts a

nd

In

jure

dA

ccid

en

ts a

nd

In

jure

d

Fa

taliti

es

Fa

taliti

es

Fa

taliti

es

Fa

taliti

es

Target Target Target Target of 2015 of 2015 of 2015 of 2015 ⇒⇒⇒⇒3000300030003000

TargetTargetTargetTarget of 2018of 2018of 2018of 2018⇒⇒⇒⇒2500250025002500

FatalitiesFatalitiesFatalitiesFatalities

4,3734,3734,3734,373

AccidentsAccidentsAccidentsAccidents

629,021629,021629,021629,021

Government Government Government Government TargetTargetTargetTargetssss

of of of of fatality reductionfatality reductionfatality reductionfatality reduction

InjuredInjuredInjuredInjured

781,494781,494781,494781,494

Vehicle FleetVehicle FleetVehicle FleetVehicle Fleet

74,434,93674,434,93674,434,93674,434,936

3

Traffic Accident in Japan

2002

N = 8,396

2013

N = 4,373

41.2%

3,463

9.3%

777

8.6%

726

11.9%

997

28.8%

2,41632.4%

1,415

6.7%

295

13.7%

600

10.6%

465

36.2%

1,584

2013 National Police Agency

Vehicle occupants

Motorcyclists

Motorized bicycle

Bicyclists

Pedestrians

Other

■Comparison of Fatalities in 2002 and 2013

4

2013 ITARDA

N = 61,001

N = 1,584

Injured

Fatal

0% 20% 40% 60% 80% 100%

Head,Face Neck Thorax,Back

Arm Abdomen Pelvis

Leg other Whole body

8.0%

5.1%

20.2%

1.0%

11.6% 33.5%

56.3% 3.7% 16.8%

4.2%

10.1%

1.5%

7.4%

20.7%20.7%20.7%20.7%

Injured Body Regions in Pedestrian

5

Hip

2012 ITARDA

Fatal

N = 64,128

N = 1,634

0% 20% 40% 60% 80% 100%

< 6 7 - 15 16 - 39 40 - 64 65 - 74 75 <

5.5%

11.9%

24.1% 29.1% 13.6% 15.7%

47.6%20.3%21.7%7.8%

< 1.5%

Age in Pedestrian

Injured

6

History of Regulation and JNCAP

Regulation 2003 2004 2005 2006 –

20102011 2012 2013 2014

Head

Leg

JNCAP 2003 2004 2005 2006 –

20102011 2012 2013 2014

Head

Leg

EEC 2003/102

BaseUN-R127 Base

EEC 2003/102

BaseUN-R127 Base

7

31 1

10

44

2 1 1 1

5

13

6 11

8 7

3 4 4

1

1

64

6 6

77 7

9 8

1 1

13

12 1

3 3

0

2

4

6

8

10

12

14

16

18

20

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

Level 1 Level 2 Level 3 Level 4 Level 5

台数

台数

台数台数

11

1212

12

0

2

4

6

8

10

12

14

16

18

20

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

Level 1 Level 2 Level 3 Level 4 Level 5

台数

台数

台数台数

Head Protection Performance Leg Protection Performance

Introduction

of regulation

Trend of J-NCAP Pedestrian Protection Test Results

Introduction

of regulation

Nu

mb

er

of

test

ed

ve

hic

les

Nu

mb

er

of

test

ed

ve

hic

les

8

Main Body Regions Injured among Fatal Pedestrians

1999 and 2009 in Real-world Accidents

60

40

20

0

Sedan Light

passenger

car

Mini

van

Box

van

Light

cargo

van

Dis

trib

uti

on

(%

)

Head

80

100

< Year 1999 >

63

8

7

22

n=940 n=165 n=68 n=149 n=246

60

12

7

21

68

12

4

16

64

8

6

21

68

15

3

14

60

40

20

0 D

istr

ibu

tio

n (

%)

80

100

< Year 2009 >

56

13

12

19

n=459 n=139 n=42 n=279 n=170

43

19

10

27

62

12

12

14

51

19

14

15

58

16

9

16

Sedan Light

passenger

car

Mini

van

Box

van

Light

cargo

van

Chest

Hip

Others

9

To reduce the fatality, a reduction of impact velocity is a key.

Head Protection Test

Detection

Auto Alert & Braking(1) Near miss incidents

AEB for Pedestrian Detection is effective.

4,373

Pedestrian1,584

36%

Motor

cycle11% Moped

7%

Bicycle60014%

32%

Car

1,415

Fatal Accident in Japan

Year 2013

(2) The effect of impact velocity reduction

(3) The current performance of AEB

Technical Countermeasure for Pedestrian Protection

10

Analysis of Near-miss Incidents Data

11

• NTSEL: 163 car-to-pedestrian near-miss incident data were

analyzed in the present study.

Near-miss Incident Data from J-SAE

Possibly recorded for 15 seconds

Sudden brake

10 seconds 5 seconds

cameraRecorder installed

When a taxi driver brakes with high

deceleration, the information are

recorded for 15 seconds.

• Data: Driving recorders are installed in taxis

2005 to 2009: 38,000 incidents as Near-miss data

105 taxis in Tokyo

20 taxis in Shizuoka

• Forward motion picture

• Acceleration

• Brake signal

• Car traveling velocity

12

Near-miss Incident (walking)

13

14

Near-miss Incident (running)

Macro Data from ITARDA

Injury level

Macro Data: All recorded accident data reported to police in Japan

Fatal, Serious, Minor

Weather Sunny, Cloudy, Rain, Foggy, Snow

Category Vehicle-pedestrian, Vehicle-vehicle,

Halt, up to 10km/h, up to 20km/h

Most severely Overall, Head, Face, Neck, Chest,

Cause e.g., Steering, Glass, Ejection from a car

injured part Abdomen, Waist, Arm, Leg

Vehicle by itself

Vehicle travel vel.

Road circumstance Straight road, Intersection

15

Fatal SeriousMinor

0102030405060708090

100

0-10

11-2

021

-30

31-4

041

-50

51-6

061

-70

71-8

081

-90

91-1

00C

um

ula

tiv

e P

erc

en

tag

e (

%)

（（（（n=1,721））））（（（（n=10,257））））（（（（n=61,010））））Near-miss（（（（n=163））））

Real world accident: 2008 Near-miss incident

FatalSerious

Minor

Near-

miss

The car travel velocity of near-miss incident was similar to

that of serious accident.

Car traveling velocity (km/h)Matsui et al. ESV 2011

Similarity of Car Traveling Velocity

between Accident and Near-miss Incident

16

17

Moving Directions Between Vehicle and Pedestrian

C

C

C

C

D

D

D

D

B

B Cross in front of the moving carA

Go to the same direction as the moving car

A A

B

B B

(1) On a straight road (2) At an intersection

18

Fatal Near-miss Fatal Near-miss

Daytime Nighttime

100

80

60

40

20

0

accident incident accident incident

Similarity of Moving Directions

between Accident and Near-missR

ati

o (

%)

A

B

C

D

38% 39%46%

38%

29%35%

32%30%

67% 74% 78% 68%

• Similarities are observed between accidents and near-miss incidents.

• We could estimate accident situations from near-miss incident data.

About 70% pedestrians were crossing roads in front of the forward

moving cars.

Cross in

front of the

moving car

Go to the same

direction as the

moving car

B

D

A

C

Matsui et al. ESV 2011

Hatching:

intersection

21%18%

16% 5%

19% 31%

19

Classification of Near-miss

Incidents

V

v

L

L Vehicle TTC =

V

LdPedestrian TTV =

v

Ld

Vehicle time to collision

(Vehicle TTC)

Pedestrian time to vehicle

(Pedestrian TTV)

Matsui et al. Traffic Injury

Prevention 2013Matsui et al. ESV 2011

Focused object

Time

Definition

Vehicle Pedestrian

Reference

Definition of Time in Vehicle TTC and Pedestrian TTV

20

21

LL

Vehicle TTCV

The worst situation was assumed that a car was

moving toward a pedestrian without braking due to a

car driver’s carelessness.

At a moment when a

pedestrian appeared

initially in front of a

car in the video frame

A drive recorder can capture forward motion pictures.

Assumption

V

=

101 near-miss incident data: Pedestrians were crossing roads.

Vehicle TTC from Near-miss Incident Data

Unobstructed From behind From behinda moving vehiclea building a parked vehicle

From behind

recorder

installingCar

drive

(1) (2) (3) (4)

view

Classified Four Pedestrians Stepping in Patterns to

Find Out a Severe Condition using Vehicle TTC

22

(n=55) (n=8) (n=10)

Avg

Ve

hic

le T

TC

(se

c)

Avg

forw

ard

dis

tan

ce o

f a

ca

r

Avg

Ve

loci

ty o

f a

ca

r (k

m/h

)

16.2m

8.6m 8.8m

11.1m

30.3km/h

24.3km/h25.8km/h

32.9km/h

2.0sec

1.4sec 1.3sec1.2sec

(n=28)

35

30

25

20

15

10

5

0

2.5

2.0

1.5

1.0

0.5

0 (1) (2) (3) (4)

& a

pe

de

stri

an

(m

)

Average Vehicle TTC

Average forward distance of a car & a pedestrian

Average car traveling velocity

the longest in (1) from unobstructed view

the shortest in (4) from behind a moving vehicle The vehicle TTC

Results of Vehicle TTC in the Four Stepping

in Patterns

23

V

v

L

L Vehicle TTC =

V

LdPedestrian TTV =

v

Ld

Vehicle time to collision

(Vehicle TTC)

Pedestrian time to vehicle

(Pedestrian TTV)

Focused object

Time

Definition

Vehicle Pedestrian

Matsui et al. Traffic Injury

Prevention 2013Matsui et al. ESV 2011Reference

Definition of Time in Vehicle TTC and Pedestrian TTV

24

The worst situation was assumed that a pedestrian

was moving toward a forward moving car line.

At a moment when a

pedestrian appeared

initially in front of a car

in the video frame

Assumption

Ld

v

Ld Pedestrian TTV

v=

Calculation of Pedestrian TTV (Time to Vehicle)

25

Results of Pedestrian TTV in the Four Stepping

in Patterns

Unobstructed view From behind a From behind a parkedFrom behind a moving

Avg

pe

de

stri

an

TT

V (

sec)

Avg

late

ral

dis

tan

ce (

m)

Avg

wa

lkin

g s

pe

ed

(m

/s)

1.8sec

1.1sec 1.1sec0.8sec

3.2m

2.2m 1.8m2.0m

1.8m/s2.1m/s

1.8m/s

2.6m/s

Average pedestrian TTV Average lateral distance of a car & a pedestrian

2.5

2.0

1.5

1.0

0.5

0

2.5

2.0

1.5

1.0

0.5

0

3.0

3.5

(1) (2) (3) (4)

Average walking speed

the longest in (1) from unobstructed view

the shortest in (2) from behind a building The pedestrian TTV

(n=55) (n=8) (n=10)(n=28)

26

2.0s

1.3s1.2s

1.4s

Average vehicle TTC Average pedestrian TTV

1.8s

0.8s1.1s 1.1s

Av

gT

TC

, T

TV

(se

c)2.5

2.0

1.5

1.0

0.5

0*Matsui et al. ESV 2013

(1) (2) (3) (4)

Each of vehicle TTC and pedestrian TTV was similar

in the 4 classified patterns.

(n=55) (n=8) (n=10)(n=28)

Comparison between Vehicle TTC and Pedestrian TTV

in 4 Classified Stepping in Patterns

27

Pedestrian Injuries and

Collision Speeds

28

29

Flow for Relations of Fatality Risks and

Impact Velocities

Fatality

risk

Travel

velocity

MACRO

DATA

micro

dataImpact

velocity

Fatality

risk

Impact

velocity

Matsui, Oikawa and Ando STAPP Jnl 2013

y = 0.9484x

R2 = 0.8189

y = 0.8929x

R2 = 0.7255y = 0.7496x

R2 = 0.6312

Minor injury Serious injury Fatal

n = 50 n = 48 n = 25

y = 0.75 x y = 0.89 x y = 0.95 x

Travel velocity (km/h)

100

80

60

40

20

0

Imp

act

ve

loci

ty

0 20 40 60 80 100

Linear Regression Coefficient

Mean

SD

P-value

0.03

0.004*

0.03 0.03 0.03

0.75 0.89 0.89 0.95

Minor Serious FatalSerious

0.180

(km/h)

Travel velocity (km/h)

100

80

60

40

20

0

Imp

act

ve

loci

ty

0 20 40 60 80 100

(km/h)

Travel velocity (km/h)

100

80

60

40

20

0

Imp

act

ve

loci

ty

0 20 40 60 80 100

(km/h)

Number 50 48 48 25

Injury level

Combined Results between Travel Velocity and

Impact Velocity for Sedan - Micro Data

30

(d) Light

passenger

car

26%

(c) Box van 4%

(e) Light

cargo van

16% (a)

Sedan

41%

(b) Mini

van

13%

1,089

Pedestrian fatal accidents in 2009 in Japan

(a) Sedan (b) Mini van (c) Box van

(d) Light passenger

car

Ordinary

automobile

Light

automobile

(e) Light cargo van

Engine displacement

≤≤≤≤ 660 cc

Category Vehicle types

Vehicle Types

31

Impact velocity

(estimated)

0 10 20 30 40 50 60 70 80

Vehicle velocity (km/h)

100

80

60

40

20

0Fata

lity

ris

k (

%)

(a) Sedan (b) Mini van (c) Box van

(d) Light passenger car (e) Light cargo van

0 10 20 30 40 50 60 70 80

Vehicle velocity (km/h)

100

80

60

40

20

00 10 20 30 40 50 60 70 80

Vehicle velocity (km/h)

100

80

60

40

20

0

0 10 20 30 40 50 60 70 80

Vehicle velocity (km/h)

100

80

60

40

20

00 10 20 30 40 50 60 70 80

Vehicle velocity (km/h)

100

80

60

40

20

0

Fata

lity

ris

k (

%)

Fata

lity

ris

k (

%)

Fata

lity

ris

k (

%)

Fata

lity

ris

k (

%)

Travel velocity

(macro data) 0.95

In case of fatal, driver did not brake enough so speed did not

reduce drastically.

Fatality Risk as Functions of Travel Velocity/ Impact Velocities

32

Matsui, Oikawa and Ando STAPP Jnl 2013 33

≤≤≤≤ 30 km/h: The fatality risks are less than or equal to 5%.

For the five types of vehicles,

We would like to propose the specification of CDMBS: ≤≤≤≤ 30 km/h

Fatality Risk as Functions of Impact Velocities

Sedan Mini van Box vanLight

passengercar

Lightcargo van

10 0 0 1 0 120 0 1 4 1 230 3 3 5 2 540 9 10 11 9 1250 22 30 31 22 2560 38 42 - 38 -

Vehicleimpactvelocity(km/h)

Fatality (%)

Performance of AEB for

Pedestrian Protection

34

Experiments for AEB Performance

Collision

case

Avoidancecase

Sensor

Frame material: FRP( Fiber Reinforced Plastics)

35

Test Conditions

Items Conditions

Test vehicle A, B, C

Detection sensor(A) (C) laser, camera and radar

(B) stereo camera

Test speed 5km/h 〜60 km/h (interval 5km/h)

Environment day, night

Surface dry, wet

Dummy position vehicle center , offsets

Dummy orientation front, side

Dummy color black, white, gray

Standard condition36

Scattered Results

0

10

20

30

40

50

60

70

0 10 20 30 40 50 60 70

Co

llis

ion

sp

ee

d (

km

/h)

Vehicle speed (km/h)

Unstable

37

38

Collision

Avoid Collision

AEBVehicle speed

Binomial Test Results

)exp(1

)exp()|1(

10

10

x

xxcp

ββββ

θ +++==

avoidc

collisionc

:0

:1

),( 10

=== ββθ

39

AEB Performance

by Binomial Logistic Regression

0

0.2

0.4

0.6

0.8

1

0 10 20 30 40 50 60 70

Pro

babi

lity

of c

ollis

ion

Vehicle speed (km/h)

Vehicle A

Vehicle B

40

Pedestrian/vehicle fatalities (2006-2010)

0

500

1.000

1.500

2.000

2.500

0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-

Ped

estr

ian

fata

litie

s

Hazard recognition speed(km/h)

fatality

Reduction of Fatality by AEB

0

500

1.000

1.500

2.000

2.500

0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-

Pe

de

stri

an

fa

tali

tie

s

Vehicle speed (km/h)

Reduction by Vehicle A

AEBFatalities

41

Three Groups of Test Results

Severe Collision

Minor Collision Threshold

Avoid Collision

AEBVehicle

speed

42

Total Risk of Severe Injury of Pedestrian

Total Severe Injury Risk: P (head and chest)

P (head or chest) = P (head)+P (chest)- P(head)*P (chest)

Collision velocity 20 km/h 30 km/h 40 km/h 50 km/h

(1) Medium Sedan 2% 5% 16% 35%

(2) Minicar 2% 4% 5% 9%

(3) SUV 2% 4% 8% 95%

43

Risk Distribution (Vehicle B)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

10 20 30 40 50 60

Pro

babi

lity

Vehicle speed (km/h)

Avoidance Minor collision Severe collision

44

AEB Effects on the Distribution of

Pedestrian Fatalities (Vehicle B)

0

500

1,000

1,500

2,000

2,500

0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-

Num

ber

of p

edes

tria

n

Vehicle speed (km/h)

Avoidance

Minor collision

Severe collision

45

46

47

Thank you

48

Fidelity Test : Radar Reflection

• Human • Dummy

49

Deterministic vs. Probabilistic

Caution!

In some objects or environment, the pre-crash brake

cannot recognize objects or does not work.

Drive carefully, without relying on the pre-crash brake

too much.