Komplexitätsreduktion und Beschleunigung: Das … 1000000 2000000 3000000 4000000 5000000 6000000...

Komplexitätsreduktion und

Beschleunigung:

Das offene Fahrerassistenz-Framework

Dr. Michael Reichel 9. November 2016

Komplexitätsreduktion und Beschleunigung: Das offene Fahrerassistenz-Framework

An ACC Developer‘s World in 2002.

© Elektrobit (EB) 2016

Radar

Target vehicle candidates, their velocity / acceleration

Steering Wheel Angle

Target vehicle selection

Motor Speed Control

Ego vehicle speed control

Head Unit

System activation, status communication

Items to specify: 4.

2


A Driverless Car Developer‘s World in 2016.


Radar

Camera

LIDAR

Sonar

Steering Wheel Sensors

Wheel Speeds

IMU / Gyro

Global Position

Maps

Motor Speed Control

Brake Pressure

Steering Angle Control

Regenerative Braking

Static obstacles

Lanes

Moving objects

Vehicle ego motion

Signs

Trajectory planning &

control

Fail-safe / -degraded

mechanisms

Head Unit

Side Tasks

Functional safety

Redundancies

Items to specify: 24.

0

5

10

15

20

25

30

2002 2004 2007 2010 2013 2015

Items to specify

1999: Mercedes S-Class Distronic

2002: VW Phaeton ACC

2005: Mercedes S-Class Distronic plus

2010: Audi A8 GPS-guided ACC

2006: Audi Q7 ACC plus / AEB

2013: Mercedes S-Class Dist.+ / Steering Assistant

2015: Audi Q7 Traffic Jam Assistant

+ brakes + sensor fusion + GPS maps + steering + automatic steering

3

0

1000000

2000000

3000000

4000000

5000000

6000000

7000000

8000000

9000000

1 3 5 7 9 11 13 15 17 19 21 23#

Inte

ract

ion

s

# Components

0

50

100

150

200

250

300

1 2 3 4 5 6 7 8 9 101112131415161718192021222324

# In

tera

ctio

ns

# Components


How Many Interactions Can We Expect For n Components?


𝑛(𝑛 − 1)

2

24 components: 276 interactions

2𝑛−1 − 1 24 components: 8.4 million interactions

Best case: Every specification gets written

or checked only once

Worst case: All specifications must be re-examined

after one is changed

4


So What Does Help?


Radar

Camera

LIDAR

Sonar

Steering Angle

Wheel Speeds

IMU / Gyro Global

Position

Maps

Automated Driving

Automated Emergency Brake

Automated Valet Parking

Interactions: 𝒏 ×𝒎 for 𝒏 sensors, 𝒎 functions

5

Sensor Data Fusion


So What Does Help?


Radar

Camera

LIDAR

Sonar

Steering Wheel Sensors

Wheel Speeds

IMU/ Gyro

Global Position

Maps

Static obstacles

Lanes

Moving objects

Vehicle ego

motion

Signs

Interactions: 𝒏 +𝒎+ 𝐤 for 𝒏 sensors, 𝒎 functions, 𝒌 abstraction components

Automated Driving



6


From Exponentiality to Linearity.


Automated Driving



Sensor Data Fusion: From echos to objects and free space

Motion Management: From brake pressure to trajectory control

HMI Management: From buttons and LEDs to user interactions

Well-defined interfaces

Well-defined interfaces

7

Automated Driving


Functional Architectures

Can we make complicated behavior even simpler?


Drive in your own lane!

Change lanes if you must!

Don‘t crash! (for heaven‘s sake!)

Come to a safe stop if necessary!

8


Possible Approach: Behavior-Based Driving

© Elektrobit (EB) 2016 10

Applicability

Desire

Risk

Comfort

HMI Control

Trajectory Plan


Applicability

Desire

Risk

Comfort

HMI Control

Trajectory Plan

Driving in Own Lane

Applicability

Desire

Risk

Comfort

HMI Control

Trajectory Plan

Automated Lane Change

Applicability

Desire

Risk

Comfort

HMI Control

Trajectory Plan


Applicability

Desire

Risk

Comfort

HMI Control

Trajectory Plan

Safe State Handling

Situative Function Arbitration

Behavior Arbitrator

Ruleset for Automated Driving

Ruleset for Manual Driving

Behavior Coordinator


Architecture Is Key To Managing Complexity.


Introducing robinos. Environment modeling and interpretation • vehicle localization in relative and absolute coordinates • environment model – objects, free space, roads, lanes, intersections, signs • interpretation of typical scenarios, recognition of a-typical scenarios

Interfaces for • interoceptive sensors – wheel ticks, steering angle, accelerometers / gyros • „smart“ environment sensors – point clouds, object lists • ADASISv3 for map, SENSORIS for cloud

Extensible function framework • rule-based coordination of functions • clear upgrade path from NCAP to HAD by adding functions • complex behavior emerges from coordinated elementary behavior • guaranteed testability / verifyability

Coordination of vehicle control and HMI • braking, steering, or full trajectory control • safe communication of vehicle state • safe hand-over / take-over management

Interfaces for • kinematic vehicle

components • instrument cluster • infotainment display


A Modular Framework for Automated Driving.


Veh

icle

Ab

stra

ctio

n -

Sen

sors

Veh

icle

Ab

stra

ctio

n -

Act

uat

ors

Safety and Error Management

Fun

ctio

n S

pec

ific

Vie

ws

HD Positioning

Object Fusion

Grid Fusion

Road and Lane Fusion

Vehicle Database

Sensor

Data Fusion

Trajectory Control

Longitudinal Control

Lateral Control

Motion Management

HMI Management

Behavior

Behavior

Behavior

Situative Behavior Arbitration

Situation Analysis

Situation Analysis

Situation Analysis

Path Planning

Path Planning

Path Planning


Fun

ctio

n S

pec

ific

Vie

ws

HD Positioning

Object Fusion

Grid Fusion

Road and Lane Fusion

Vehicle Database

Sensor

Data Fusion

Trajectory Control


Lateral Control

Motion Management

HMI Management

Behavior

Behavior

Behavior


Situation Analysis

Situation Analysis

Situation Analysis

Path Planning

Path Planning

Path Planning V

ehic

le A

bst

ract

ion

- A

ctu

ato

rs

Veh

icle

Ab

stra

ctio

n -

Sen

sors

Integrated safety concept • system health monitoring and diagnosis • safe-state triggering • options for redundant environment model

and functions (e.g. minimal risk manoeuvers)

Fun

ctio

n S

pe

cifi

c V

iew

s

Ve

hic

le A

bst

ract

ion

- S

en

sors

Ve

hic

le A

bst

ract

ion

- A

ctu

ato

rs

Trajectory Control


Lateral Control

Motion Management


Behavior

Behavior

Behavior


Situation Analysis

Situation Analysis

Situation Analysis

Path Planning

Path Planning

Path Planning V

eh

icle

Ab

stra

ctio

n -

Se

nso

rs

Fun

ctio

n S

pe

cifi

c V

iew

s

HD Positioning

Object Fusion

Grid Fusion

Road/Lane Fusion

Vehicle Database

Sensor Data Fusion

Behavior

Behavior

Behavior


Situation Analysis

Situation Analysis

Situation Analysis

Path Planning

Path Planning

Path Planning

Trajectory Control


Lateral Control

Motion Management

HMI Management

Ve

hic

le A

bst

ract

ion

- A

ctu

ato

rs



robinos Key Concepts: Standardized Interfaces


Every software component has • scalable, documented and standardized interfaces to other components • exchangeable interfaces to the base system / OS • a pre-industrialized algorithm core


Common best practices accelerate testing


Fun

ctio

n S

pe

cifi

c V

iew

s

Ve

hic

le A

bst

ract

ion

- S

en

sors

Ve

hic

le A

bst

ract

ion

- A

ctu

ato

rs

Trajectory Control


Lateral Control

Motion Management


Behavior

Behavior

Behavior


Situation Analysis

Situation Analysis

Situation Analysis

Path Planning

Path Planning

Path Planning V

eh

icle

Ab

stra

ctio

n -

Se

nso

rs

Fun

ctio

n S

pe

cifi

c V

iew

s

HD Positioning

Object Fusion

Grid Fusion

Road/Lane Fusion

Vehicle Database

Sensor Data Fusion

Behavior

Behavior

Behavior


Situation Analysis

Situation Analysis

Situation Analysis

Path Planning

Path Planning

Path Planning

Trajectory Control


Lateral Control

Motion Management

HMI Management

Ve

hic

le A

bst

ract

ion

- A

ctu

ato

rs



robinos Key Concepts: Smart Redundancy


Fun

ctio

n S

pe

cifi

c V

iew

s

Ve

hic

le A

bst

ract

ion

- S

en

sors

Ve

hic

le A

bst

ract

ion

- A

ctu

ato

rs

Trajectory Control


Lateral Control

Motion Management


Behavior

Behavior

Behavior


Situation Analysis

Situation Analysis

Situation Analysis

Path Planning

Path Planning

Path Planning V

eh

icle

Ab

stra

ctio

n -

Se

nso

rs

Fun

ctio

n S

pe

cifi

c V

iew

s

HD Positioning

Object Fusion

Grid Fusion

Road/Lane Fusion

Vehicle Database

Sensor Data Fusion

Behavior

Behavior

Behavior


Situation Analysis

Situation Analysis

Situation Analysis

Path Planning

Path Planning

Path Planning

Trajectory Control


Lateral Control

Motion Management

HMI Management

Ve

hic

le A

bst

ract

ion

- A

ctu

ato

rs


15

HMI Management

HMI Ctrl

Traj. Plan

Object Fusion

Grid Fusion

Time to collision

• Model-based (EKF) • Classified, dynamic objects • ASIL possible

• Model-free (evidence-based) • Dynamic freespace • ASIL possible

Time to collision

Safety Decision

Single point of failure eliminated

configurable sensor mapping


A Modular Framework enables you to...


Upgrade across models / Upgrade over time

NCAP

HAD

Ensure differentiation, shorten time to market

EB modules 3rd-party modules Your modules

Map onto concrete ECU architecture


Best Practice Leverage: Project, Product Family, Industry


Project-Level Architecture Best Practices: Reduce complexity

€

Product Family Architecture Best Practices: Re-use specified, developed, industrialized and tested components

€€

Industry-Wide Architecture Best Practices Peer review of functionalities, safety and security mechanisms Lower entry into HAD development No vendor lock-in

€€€


Open robinos and EB robinos


Open robinos

• specifies a reference platform for automated driving up to Level 5 (SAE) • architecture • interfaces • data flow • control mechanisms • software modules • functional safety aspects

• freely available and

licensed as Creative Commons

• Available for download today

EB robinos

• implements the open robinos specification

• provides software modules • for prototyping

in EB Assist ADTF • for rapid embedding

on AUTOSAR / DrivePX • for production

on vehicle ECU

• developed, tested,

verified according to functional safety standards

18

Download the open robinos specification

www.open-robinos.com

working group members welcome!


Q4/2016 – More EB robinos components


Thank you!


Industry-Wide Architecture Best Practices Peer review of functionalities, safety and security mechanisms Lower entry into HAD development No vendor lock-in

€€€

www.eb-robinos.com www.try-eb-robinos.com

[email protected]

http://www.eb-robinos.com/



http://www.try-eb-robinos.com/





mailto:[email protected]

Komplexitätsreduktion und Beschleunigung: Das … 1000000 2000000 3000000 4000000 5000000 6000000...

Documents

Transcript of Komplexitätsreduktion und Beschleunigung: Das … 1000000 2000000 3000000 4000000 5000000 6000000...