adoption of connected automated vehicles for traffic … · denso international america, inc....
TRANSCRIPT
Roger BergVice PresidentNorth America Research & DevelopmentDENSO INTERNATIONAL AMERICA, INC.
Adoption of Connected Automated Vehicles for Traffic Safety
Global ADAS Regulatory Trend
CIB (Car)
AEBS&LDW for truck
CIB (Car & Ped)
L Kids & Transportation
NN
AEBS (Car) / LDWISAAEBS (Ped)
14 15 16 17 18~~13
Rear cameraN
L LKA
AEBS-PDLDW
AEBS for truck
AEBSTFCWT
L N TLaw (Mandatory) NCAP Top Safety Pick (IIHS)
L
N NN
N
N
N
AEBS: Advanced Emergency Brake SystemCIB: Crash Imminent BrakingFCW: Forward Collision WarningLKA: Lane Keep AssistLDW: Lane Departure WarningISA: Intelligent Speed Adaptation
cartalk.com
http://www.cvel.clemson.edu/
http://virtual.vtt.fi/
Global ADAS Use Case Trend
ComfortConvenience
Collision Avoid
Road Departure Prevent
Surround Monitor
Appl
icat
ion
Night Assist
Low Speed AEBS
LDW
FSRA
LKA
ESPS
PCS
AHB
FLAD(On demand lane change, fork/merge
management )
BSW/CVW/RCTA/RPCS/FCTA
LDC RDP
New ADB
TSR
FSRA/ACC w/ stationary vehicle
detectionFLAD
within lane(FSRA+LKA+TSR)
ADB
AEBS w/ steering support
AEBS (Pedestrian)Night perf. enhancement
AEBS
AEBS(Pedestrian)
CS
14 15 16 17 18 1913
Camera
Sonar
Laser
77GHZ Radar
FSRA : Full Speed Range ACCLKA: Lane Keep AssistLDW: Lane Departure WarningTSR: Traffic Sign RecognitionFLAD: Freeway Limited Autonomous DrivingPCS: Pre-Crash SafetyAEBS: Advanced Emergency Brake SystemRDP: Road Departure Prevention
CS: Clearance SonarICS: Intelligent CS APA: Advanced Parking AssistBSW: Blind Spot WarningCVW: Closing Vehicle WarningRCTA: Rear Cross Traffic AlertRPCS: Rear PCSFCTA: Front Cross Traffic Alert
AHB: Auto High BeamADB: Adaptive Driving BeamLDC: Lane Departure Controller
9/22
Options to Choose
vision sensor (camera)LIDAR mm wave RADAR sonar
telematics control unit (cellular radio)
FCW, LKA software
In Production
Next Generation
V2x radio
LIDAR+ camera
Next gen LIDAR
surround
Next Gen front LIDAR
pedestrian detection
GPSLocator
mini stereo camera
3D flashLIDAR
10/22
Sensors are Imperfect
On Board Sensor CapabilityLIDAR Camera RADAR V2X &
Cellular*Map &
Location
traffic objects
Car / Trucks
Motorcycle
Bicycle
Pedestrian
Animal
Debris
infrastructure traffic constraints
Road marking
Traffic signs
Traffic lights
path /road structure
Curve
Guard rails
Wall/obstacle
RobustnessRain / Snow
Fog
Night
Det
ectio
n Pe
rfor
man
ce
dete
ctio
n &
rang
ing
dete
ctio
n &
rang
ing
mon
itorin
g &
cla
ssifi
catio
n
info
rmat
ion
info
rmat
ion
Dynamic Static
*depends on coverage
11/22
Combination of HW/SW ≈ CPS
Sensing Vehicle Driverautonomouscooperativespeed
yaw
throttle
brake
motion HMIglobalmap
• road• landmark• traffic rule ...
sensor fusion processing
sensorcomponents
signal process &detection
monitoring &recognition surroundings recognition (local map)
power train brake steer chassis
ADAS & AD applications
arbitration &distribution
actuation
longitudinal acc. arbitration lateral acc./yaw rate arbitration
HMI
drive mode transition
distribution (accel. / decel.) distribution (steer / yaw)
ADAS App. #1 AD App. #2AD App. #1 ・・・・・・
open I/F
open I/F
LIDAR camera
radarsonarV2Xradio
cellularradio
GPSlocator
drivermonitor
biosensor
driver monitoring
open I/F
Sensor System
steer
open I/Fintervention from vehicle safety
gyro
ADAS App. #2
Not one company can do it all No longer is a vehicle its own entity, separate from
infrastructure and other road users We can’t necessarily predict how connected and automated
drivers interact with “regular” vehicles of all types
Requires collaboration between industry, academia, and governments to be effective.
Cooperative Systems Have Taught Us
With connectivity comes…responsibility!
Infiltration end point is an ECU running a critical vehicle control function
Vehicle
ECU
ECU ECU ECU ・・・
uC
RemoteAgent
Architecture 1. remote
2. in-vehicle network
3. ECU direct
Why does automotive need to study this?- Big impact of neglect for entire AD vehicle industry- Entire supply chain may be liable
But what should we do?- Better understand today’s threat, impact, resolution- Forecast and adapt to future UNKNOWN threats
18 /32
Conclusion
A continuum of ADAS is evolving from no assistance at all to the driver to full control by cyber systems.
A suite of devices with multi spectral sensing (e.g., RADAR, LIDAR, camera, ultrasonic) is likely to be necessary but not sufficient for all scenarios.
One approach is to evolve the idea of cooperative sensors (e.g. V2X, cellular WAN, GPS/GNSS) via connectivity to enhance robustness.
Stakeholders must collaborate and cooperate across industry, academia and government for safe and comfortable connected automation.
Understand the importance of security across the supply chain to help ensure trust in the CPS which assumes control of the vehicle.
19/22
Roger BergVice President North America R&DDENSO INTERNATIONAL AMERICA, [email protected](760) 597 - 7404
2015 202X
Front Sensing
Radar(Front side & Rear Side)
CameraRadar(Front)
Function & O
bjectRecognition
Recognition D
omain
2015 202X
・Cut in vehicle ・Safe headway
’16 NCAP ’18 NCAP
・Moving vehicle・Stopped vehicle
AEB Crossing(Cyclist)
Single-lane driving Multiple-lane change
AEB Pedestrian
Full-speed-range ACC
Split & Merge Emergency stop(Driver/system fail)
AEB Pedestrian (night)
Highway
OrdinaryRoad
・Pedestrian (day) ・Pedestrian (night)・Cyclist
※Blue means recognition object
・Merging vehicles・Road edge
Driver Assistance
Advanced driver assistance(SAE:LV2)
Fully automated driving(SAE:LV4/5)
Advanced driver assistance(SAE:LV3)
2020
Sensing technology+ Surround information linkage ( V2X , ADAS map )
Driver assistance
Front + Side + Rear Sensing
Recognition Objects & Domain expansion, Look-ahead of local information
Active safetyCooperative Safety Assistance
(e.g., V2X communication)
・Obscured vehicle ・Crossing vehicle・Oncoming vehicle
AEB Intersection Vehicle
ADAS – Automation Roadmap
http://blogmedia.dealerfire.com/wp-content/uploads/sites/193/2015/02/Following-Distance-in-Fog.jpg
https://ledouxinsurance.files.wordpress.com/2011/10/driving-in-rain1.jpg http://il5.picdn.net/shutterstock/videos/9994577/thumb/1.jpg?i10c=img.resize(height:160)
https://upload.wikimedia.org/wikipedia/commons/b/b9/Cars_in_snowstorm.JPG
AD in Unfavorable Conditions