autonomous longitudinal control with driving advices Dipl. Ing. Bjrn - PowerPoint PPT Presentation

Continuous testing over the entire development process of autonomous longitudinal control with driving advices Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology System Requirements Analysis Test Integration System Design Test Subsystem Functional Design Test M Module Module Design Test Module Implementation KIT - University of the State of Baden-Wuerttemberg and www.kit.edu National Laboratory of the Helmholtz Association

Inhalt Motivation Unit under Test Driving Advices in CarMaker Process of testing Offline Test Hardware in the loop testing Vehicle test Conclusions and results Outlook 2 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

Motivation Many tests during development of driver assistance systems Vehicle Prototype Very hardly tests expensive available Tests during whole development process Save Save time money 3 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

Unit under Test Driving Assistance system EeDA – E fficient eD rive A ssistant EeDA Automated Assisted Vehicle is following a calculated Driver gets advices to perform trajectory autonomously. Driver an efficient way of driving has only to steer. 4 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

Unit under Test EeDA Adaptive v(s) Trajectory Motor torque CAN bus data vehicle controller generation model Internal route data Route data Route data processing Driving Driving Optic and haptic feedback for CAN bus data style advices driver recognition 5 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

Driving Advices Debugging with plots or numbers only is hard  visual driving advices where implemented as Overlay in Movie using Tcl/Tk recommended behaviour for next current speed limit situation speed advice regarding next situation Distance to next situation Distance to next recommended type of next situation behaviour 6 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

Driving Advices speed limit 50 curve left curve right hilltop more possible 9 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

Driving Advices Debugging with plots or numbers only is hard  visual driving advices where implemented as Overlay in Movie using Tcl/Tk recommended behaviour for next current speed limit current speed limit situation speed advice regarding next speed advice regarding next situation situation Distance to next situation Distance to next recommended type of next situation type of next situation behaviour 10 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

Driving Advices Acceleration Sailing Braking Recuperation stage 1 Recuperation stage 2 Recuperation stage 3 11 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

V-model System Requirements Test Analysis System Integration Design Test Subsystem Functional Design Test Module Module Design Test Module Implementation 14 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

Process of testing – Offline test System Requirements Test Analysis System Integration Design Test Subsystem Functional Design Test Module Module Design Test Module Implementation 15 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

Offline test - Overview Requi reme System nts Test Analy sis Syste m Integration Desig Test n Subsy Functi stem onal Design Test Functional module test Modul Modul M e e Desig Test n Module Implementation CarMaker-Office is used Modules are build as Plugin-models with Matlab/Simulink and compiled with Realtime Workshop Route data could not be received from external source (navigation system, Car-PC, …)  method to determine route data directly from CarMaker needed 16 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

Process of testing – route data generation Requi reme System nts Test Analy sis Syste m Integration Desig Test n Subsy Functi stem onal Design Test Modul Modul M e e Desig Test n Module Implementation CarMaker C#-Code Simulink Vector of Road model route data • calculating road bend • Contains all • more than 3000 • Same format • determining needed blocks like in the car slope and features speed limit 17 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

Process of testing – Hardware in the loop test System Requirements Test Analysis System Integration Design Test Subsystem Functional Design Test Module Module Design Test Module Implementation 18 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

HIL Test - Overview Require System ments Test Analy sis Syste m Integration Desig Test n Software on ETAS ES910 rapid prototyping ECU Subsyste Functional m Design Test M Module Module Design Test CarMaker-HIL with Xpack4 Module Implementation Route Data provided by separate CarPC Engine controller in the loop Residual bus simulation of 500 signals and 4 CANs in Simulink 19 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

HIL Test - HIL test bench set-up Require System ments Test Analy sis Syste m Integration Host PC Desig Test n Subsyste Functional m Design Test M Module Module test control Design Test Module Implementation Ethernet IPG Xpack 4 vehicle model environment model driver model drivetrain model CAN CAN CAN ETAS ES910 Engine Controller CarPC EeDA software torque distribution digital road data safety functions 20 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

HIL Test – tests and test track Require System ments Test Analy sis Syste m Integration Desig Test n Basic functionality tests with IPG Testmanager Subsyste Functional m Design Test M Module Module Design Test activation/deactivation of EeDA in different scenarios Module Implementation safety functions of engine controller Import of real test track into CarMaker 21 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

Process of testing – vehicle test Requirements System Test Analysis Integration System Design Test Subsystem Functional Design Test Module Module Design Test Module Implementation 22 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

vehicle test - Overview Requirements System Analy Test sis Integr System ation Design Test Software was tested in the car using an ETAS ES910 Functio Subsystem nal Design Test Modul Module e Design Test Module Implementation Works like espected with very little application effort Driver Advices in the car 23 23.09.2016 Dipl. Ing. Björn Fath und Dipl.-Inform. Dominik Dörr Institute of vehicle system technology

autonomous longitudinal control with driving advices Dipl. Ing. Bjrn - PowerPoint PPT Presentation

Continuous testing over the entire development process of autonomous longitudinal control with driving advices Dipl. Ing. Bjrn Fath und Dipl.-Inform. Dominik Drr Institute of vehicle system technology System Requirements Analysis Test

Panel: Human Interactive Autonomous Driving Future trends Fully automated systems

Self Driving Car Self Driving Cars Auto Breaking Fully Lane Guidance Autonomous Auto Parking

Autonomous Driving: The Good The Bad and The Ugly When do you launch the product? Whos

AUTONOMOUS DRIVING IN AGRICULTURE LEADING TO AUTONOMOUS WORKSITE SOLUTIONS Dr. John F. Reid,

Comp 790-058 Lecture 10: Autonomous Driving: Control, Traffic, Predictions October 24, 2017

T echnical and Legal Challenges for Urban Autonomous Driving Seung-Woo Seo, Prof. Vehicle

Visual Scene Understanding for Autonomous Driving Raquel Urtasun University of Toronto Oct 3,

AUTONOMOUS DRIVING FRENCH NATIONAL PLAN NEW FRANCE FOR INDUSTRY (NFI) PLAN JF SENCERIN

AI and Self-Driving Cars Heechul Yun Autonomous Car

Car to car communication of autonomous driving vehicles in dangerous situations NAME: FABIAN

Autonomous driving : French policy update F-US roundtable Connected and autonomous vehicles : a

S9932: LEARNING TO BOOST S9932: LEARNING TO BOOST ROBUSTNESS FOR ROBUSTNESS FOR AUTONOMOUS

Commercialize Autonomous Driving From Lab to Production Line Dr. Xing Yuan | Baidu, Inc. Now,

DEEP LEARNING IN THE FIELD OF AUTONOMOUS DRIVING AN OUTLINE OF THE DEPLOYMENT PROCESS FOR ADAS

Overview of Autonomous Driving Sept 26, 2017 Sahil Narang University of North Carolina, Chapel

W ISE M OVE ? A research platform that mimics our autonomous driving stack. Objective:

Autonomous Motion Planning for an Automotive System Alan Kuntz Autonomous Driving in Urban

A Content-Centric Network for Autonomous Driving Swarun Kumar Lixin Shi, Stephanie Gil, Nabeel

Apply Image-to-Image Translation on Autonomous Driving Systems Testing Presented by Yilin Han,

Self-Driving or Autonomous Networks Dr. Mariam Kiran Scientific Networking Division

Learning Driving Styles for Autonomous Vehicles for Demonstration Markus Kuderer, Shilpa

Measuring the World: Designing Robust Vehicle Localization for Autonomous Driving Frank Schuster,

The Latest of Project Apollo Bringing autonomous driving to production Baidu IV Oct 2017 Apollo

Using Git Matthieu Moy Matthieu.Moy@imag.fr 2016 Matthieu Moy (Matthieu.Moy@imag.fr) Git 2016