dna for automated driving - gtc on demand · 2017. 5. 19. · dna for automated driving conclusion...
TRANSCRIPT
-
DNA for Automated Driving
Jeremy DahanMay 8th, 2017
-
© Elektrobit (EB) 2017 | Confidential
DNA for Automated Driving
2
Radar
Camera
LIDAR
Sonar
SteeringWheel Sensors
WheelSpeeds
IMU /Gyro
GlobalPosition
Maps
Motor SpeedControl
Brake Pressure
SteeringAngle Control
RegenerativeBraking
Static obstacles
Lanes
Moving objects
Vehicleego motion
Signs
Trajectoryplanning &
control
Fail-safe /-degraded
mechanisms
Head Unit
Side Tasks
Functionalsafety
Redundancies
Items to specify: 24
0
5
10
15
20
25
30
2002 2004 2007 2010 2013 2015
Items to specify
1999: Mercedes S-ClassDistronic
2002: VW PhaetonACC
2005: Mercedes S-ClassDistronic plus
2010: Audi A8GPS-guided ACC
2006: Audi Q7ACC plus / AEB
2013: Mercedes S-ClassDist.+ / Steering Assistant
2015: Audi Q7Traffic Jam Assistant
+ brakes + sensor fusion + GPS maps + steering + automatic steering
-
© Elektrobit (EB) 2017 | Confidential
Sensor Data Fusion
DNA for Automated Driving
3
Radar
Camera
LIDAR
Sonar
SteeringWheel Sensors
WheelSpeeds
IMU/Gyro
GlobalPosition
Maps
Static obstacles
Lanes
Moving objects
Vehicleego motion
Signs
Interactions: 𝒏 +𝒎+ 𝐤for 𝒏 sensors, 𝒎 functions, 𝒌 abstraction components
Automated Driving
Automated Emergency Brake
Automated Valet Parking
-
© Elektrobit (EB) 2017 | Confidential
DNA for Automated Driving
4
-
© Elektrobit (EB) 2017 | Confidential
DNA for Automated Driving
Software Framework for ADAS and Automated Driving
5
Interfaces for• Interoceptive sensors – wheel ticks,
steering angle, accelerometers / gyros• „Smart“ environment sensors – point
clouds, object lists• ADASISv2/3 for map, SENSORIS for cloud
Integrated safety concept• System health monitoring and diagnosis• Safe-state triggering• Options for redundant environment
model and functions (e.g. minimal risk
Interfaces for • Kinematic vehicle components• Instrument cluster• Infotainment display
www.open-robinos.com
Trajectory Control
Longitudinal Control
Lateral Control
Positioning
Object Fusion
Grid Fusion
Behavior
Road & Lane Fusion
Vehicle DatabaseFu
nct
ion
Sp
eci
fic
Vie
ws
Ve
hic
le A
bst
ract
ion
-Se
nso
rs
Ve
hic
le A
bst
ract
ion
-A
ctu
ato
rsMotionManagement
Safety Management
Behavior
Behavior
Situative Behavior ArbitrationSensorData Fusion
HMI Management
SituationAnalysis
SituationAnalysis
SituationAnalysis
PathPlanning
PathPlanning
PathPlanning
http://www.open-robinos.com/
-
© Elektrobit (EB) 2017 | Confidential 6
Every software component has• scalable, documented and standardized interfaces to other components• exchangeable interfaces to the base system / OS• a pre-industrialized algorithm core
DNA for Automated Driving
-
© Elektrobit (EB) 2017 | Confidential
DNA for Automated Driving
7
Map onto concrete ECU architecture
Upgrade across models / Upgrade over time
NCAP
HAD
Ensure differentiation, shorten time to market
EB modules 3rd-party modulesYour modules
-
© Elektrobit (EB) 2017 | Confidential
DNA for Automated Driving
8
Trajectory Control
Longitudinal Control
Lateral Control
Positioning
Object Fusion
Grid Fusion
Road & Lane Fusion
Vehicle Database
Fun
ctio
n S
pe
cifi
c V
iew
s
Ve
hic
le A
bst
ract
ion
-Se
nso
rs
Ve
hic
le A
bst
ract
ion
-A
ctu
ato
rsMotionManagement
Safety Management
Behavior
Behavior
Situative Behavior ArbitrationSensorData Fusion
HMI Management
SituationAnalysis
SituationAnalysis
PathPlanning
PathPlanning
Automated ValetParking
Application example: Automated Valet Parking
-
© Elektrobit (EB) 2017 | Confidential 9
DNA for Automated Driving
9
-
© Elektrobit (EB) 2017 | Confidential
Short/Mid-term Path Planning
Grid Freespace Extraction
Highway Driving
Positioning
Software framework in action
DNA for Automated Driving
10
-
© Elektrobit (EB) 2017 | Confidential
DNA for Automated Driving
Range of ego sensors are limited
•
•
Recognition algorithms are limited
Not all information needed can be derived from sensor observations
•
•
•
•
11
What about maps?
-
© Elektrobit (EB) 2017 | Confidential
DNA for Automated Driving
12
Sensor-based Learning for Predictive Driving
• Secure connection
and transfer
Collect sensor data
1
• Data aggregation• Map matching• Sensor fusion
Sensorlearning
2
• Enriched map
updates on
daily basis
IncrementalNDS Compilation
3
• Delta maps
Incremental MapUpdate Service
4
Consumerexperience
• Always up to date
electronic horizon for ADAS functions
5
2 3 4
51
-
© Elektrobit (EB) 2017 | Confidential 13
How sensor data is collected and sent
• Objects form camera recognition modules
• Create 2.5D maps of the road
• Algorithms for feature extraction:
image processing
machine learning
• Objects that can be extracted:
lane markings
lane Geometry
road boundaries
lane arrows
traffic signs
Image processing
DNA for Automated Driving
2 3 4
51
-
© Elektrobit (EB) 2017 | Confidential
How sensor data is collected and sent
DNA for Automated Driving
Top View
14
Height Map
2 3 4
51
-
© Elektrobit (EB) 2017 | Confidential
Example: Zebra crossing recognition on 2.5D maps
15
DNA for Automated Driving
2 3 4
51
-
© Elektrobit (EB) 2017 | Confidential
DNA for Automated Driving
16
How to learn from sensor data and enrich maps
• Data cleanup
Incoming data is validated regarding location and timestamp
Wrong movement profiles (e.g. ferries or trains) are discarded
• Data aggregation
Collected data is aggregated into clusters to ensure reliability
Clusters are matched on a recent map
User data is anonymized to ensure privacy
Data processing
2 3 4
51
-
© Elektrobit (EB) 2017 | Confidential
Example: Zebra crossing in the Cloud
17
DNA for Automated Driving
2 3 4
51
-
© Elektrobit (EB) 2017 | Confidential 18
Maps boost ADAS and Automated Driving
DNA for Automated Driving
-
© Elektrobit (EB) 2017 | Confidential
DNA for Automated Driving and NVIDIA
19
EB robinos
EB Assist ADTF EB tresos
DNA for automated driving
Rapid prototyping
C, C++, Model based
Rapid embedding
C, C++
Automotive gradesoftware
ECUEvaluation hardware
PC
-
© Elektrobit (EB) 2017 | Confidential
DNA for Automated Driving
Open robinos
• specifies a reference platform forautomated driving up to Level 5 (SAE)• architecture• interfaces• data flow• control mechanisms• software modules• functional safety aspects
• freely available and licensed asCreative Commons
• Available for download
EB robinos
• implements the open robinosspecification
• provides software modules• for prototyping
in EB Assist ADTF• for rapid embedding
on AUTOSAR / DRIVE PX• for production
on vehicle ECU
• developed, tested, verifiedaccording to functional safetystandards
www.try-eb-robinos.com
20
Download the open robinos specification
www.open-robinos.com
http://www.try-eb-robinos.com/mailto:[email protected]
-
© Elektrobit (EB) 2017 | Confidential
DNA for Automated Driving
ConclusionSoftware frameworks and functional
architectures help solve it.
We should talk about software, not hardware architectures.
Q4/2016 – More EB robinos componentsThe problem is not difficulty but complexity.
EB robinos is a software framework for automated driving, applicable across car lines and models – it is DNA for automated driving
21
-
Get in touch! Thank you!
www.eb-robinos.com
www.try-eb-robinos.com
Positioning (robinos, ground truth)
Automated Valet Parking
Automated Highway Driving
Electronic Horizon
https://www.facebook.com/EBAutomotiveSoftwarehttps://www.facebook.com/EBAutomotiveSoftwarehttps://twitter.com/EB_automotivehttps://twitter.com/EB_automotivehttps://www.youtube.com/user/EBAutomotiveSoftwarehttps://www.youtube.com/user/EBAutomotiveSoftwarehttps://www.linkedin.com/company/elektrobit-eb-automotivehttps://www.linkedin.com/company/elektrobit-eb-automotivehttp://www.eb-robinos.com/http://www.try-eb-robinos.com/mailto:[email protected]