Kooperative Sensorik für die Fussgängersicherheit Car2VRU-Kommunikation

Transcription

1 Kooperative Sensorik für die Fussgängersicherheit Car2VRU-Kommunikation i Wireless Communication and Information Berlin, Prof. Dr.-Ing. Dipl.-Ing. Dipl. Wirt.-Ing. Axel Sikora Department Information Technology Baden-Wuerttemberg Cooperative State University Loerrach, 1

2 ch.1 Visions & Definitions ch.2 Image Based Systems ch.3 Extensions with Cooperative Sensors ch.4 Summary & Outlook 2

3 ch.1: Vision (1) Cooperative Traffic Future traffic can be seen as a cooperative system, where devices, vehicles and infrastructures are context aware and capable of providing the needed information seamlessly and on time. ON PEDESTRIANS / CYCLISTS A wearable communication module (e.g. in rucksacks, watches, mobile phones ) ON CARS / TRUCKS A communication and localisation module A vision based sensor An on board unit for data fusion An HMI to warn the driver ON MOTORBIKES A communication and localisation module An HMI to warn the driver Source: A. Lübke, Car-to-Car Communication - Technologische Herausforderungen, VDE-Kongress 2004, , Berlin + Safety +Mobility Source: L. Andreone, A. Guarise, Vulnerable Road Users thoroughly addressed in accident prevention: the WATCH-OVER European project, ITS Environment + Infrastructure 3

4 ch.1: Vision (1) Cooperative Traffic German and EC project examples CVIS Cooperative Vehicle-Infrastructure Systems Coopers CO-OPerative SystEms sfor Intelligent Road Safety SEVECOM "SEcure VEhicle COmmunication SAFESPOT Smart Vehicles on Smart Roads GST "Global System for Telematics PreVent PReVENTive and Active Safety Applications 4

5 ch.1: Vision (2) Accident Free Driving VRUs special emphasis on Vulnerable Road Users Major element to further reduce EU fatalities half by 2010 (2000: 54k target 2010: 27k; is 2009: 34.5k) half again by

6 Ein Projekt der ch.2: sensors for VRU detection successor to NIRWARN project 6

7 ch.2: sensors for VRU detection training & performance Classification performance 1 0,95 0,9 Detection rate 0, ,8 0, ,7 0 0,05 0,1 0,15 0,2 0,25 0,3 False alarm rate Quelle: Gavrila, Daimler AG 7

8 ch.2: sensors for VRU detection image based problems Quelle: BMW Forschung und Technik 8

9 ch.1 Visions & Definitions ch.2 Image Based Systems ch.3 Extensions with Cooperative Sensors ch.4 Summary & Outlook 9

10 ch.3: Definition (1) Cooperative Sensor Network A Cooperative Sensor Network is an array of sensors interconnected dby a communications i network. Sensor data is shared between the sensors and used as input to an estimator tor to measure a distributed ted process. S S S S event/process S S S S S S S S S S S S 10

11 ch.3: Definition (2) Cooperative Sensors Cooperative Sensors are sensors that measure some input data based on their spatially distribution. A (communication) channel between the sensors is required. S S S 11

12 ch.3: Basic Discussion Cooperative Sensors conventional sensors radar laser Radar ( Lidar ) near and far infrared cameras 3D-(TOF)-camera stereo camera ultrasonic + accurate positioning of objects - mostly (LOS) needed - classification of objects by perception of sensors - limited potential for improving detection and classification - influence from weather & noise cooperative sensors transponder attached to road users sends anonym ID code position of transponder is located + positioning possible + no line-of-sight (LOS) needed + classification by ID + immanent plausibility + better SNR (1/R²) - other road users need to be cooperative - limited accuracy - depending on bandwidth, cost & time 12

13 ch.3: WATCH-OVER architecture Sensor fusion of camera and a radio based positioning system Radio tag and a reduced user interface for pedestrians and two-wheelers Cooperative relative positioning of two or more modules in range Ad-hoc network establishment tbetween radio modules 13

14 ch.3: WATCH-OVER hardware 14

15 ch.3: WATCH-OVER measurement results 150,00 100,00 50,00 dyna amic error (cm) 0, ,00-100,00-150,00 distance (m) mean upper bound(100%) lower bound (100%) 15

16 ch.3: Amulett project Supported by Bavarian Ministry of Economic Affairs, Infrastructure, Transport and Technology concentration on true cooperative sensing combination of TOF & AOA 16

17 ch.3: Amulett project Parameter / Condition Operational Range (Line-of-Sight Contact) Worse-Case Operational Range (Non-Line-of-Sight Scenarios) Range Measurement Accuracy (Line-of-Sight Contact) Range Measurement Accuracy (Non-Line-of-Sight Scenarios) Angular Measurement Accuracy (Line-of-Sight Contact) Angular Measurement Accuracy (Non-Line-of-Sight Scenarios) Measured Results for AMULETT System 0.1 to 400 m -45 to +45 deg 25 m 0.05 m 0.5 m 0.8 deg 2deg 17

18 ch.3: Amulett project 18

19 ch.3: Ko-TAG cooperative system Ko-FAS initiative for development of Cooperative Driver Assistance Systems Ko-TAG Project for Cooperative Localization Partners: BMW Forschung und Technik GmbH, München Continental Safety Engineering International GmbH, Alzenau Daimler AG, Sindelfingen Fraunhofer Institut für Integrierte Schaltungen, Erlangen Steinbeis Innovation Center Embedded Design and Networking (sizedn) Technische Universität (TU) München 19

20 ch.3: Ko-TAG cooperative system 20

21 ch.3: KoTAG Protocol Design overlaid star topology restrict to car2vrucommunication synchronization niz n of vehicles for reduced d collision i probability bilit out-of-band synchronization (GPS-/Galileo-clock, UMTS) 21

22 ch.3: KoTAG Protocol Design proposal: Multi Channel Splitted Localization (MCSL) 3 channels compatible with 5.9GHz-C2C-band for future integration combination of 10MHz: angle of arrival (AOA) measurement 30MHz: time of flight (TOF) measurement + data communication 10MHz: independent control/management channel scalability of system performance through parameterization # of transceivers 22

23 ch.3: KoTAG Design of Data Link Layer Protocol Device nextlayer 1: MLM_ANNOUNCEMENT.request Device MAC layer RF Coordinator MAC layer Coordinator next layer Tag Announcement 2: announcement request 3: MLM_ ANNOUNCEMENT.inication 4: MLM_ANNOUNCEMENT.response 7: MLM_ASSOCIATE.request 5: announcement response Vehicle Announcement 6: MLM_ANNOUNCEMENT.confirm Vehicle connection setup 8: associate request 9: ACK 10: MLM_ASSOCIATE.inication 11: MLM_ASSOCIATE.response 12: associate response 13: MLM_ASSOCIATE.confirm 14: ACK Response with address and time slot assignment 23

24 ch.3: KoTAG Simulation of Data Link Layer Protocol network simulation with OpNet Modeler close to real physical channel characteristics (data rate, bandwidth ) definition of relevant scenarios and statistics to verify proposed protocol moving nodes based on dfi definedd pathes or on random parameterization of network protocol scenario with moving node statistical evaluation 24

25 ch.4: Further Activities system design finalize simulation and parameterization mobility models hardware design p integration time synchronisation of vehicles TOF measurement system integration software design implementation of cooperative sensor protocol sensor fusion TOF & AOA communication to OBU security & safety discussions 25

26 ch.4: Thank you Contact: Prof. Dr. Axel Sikora Bd Baden-Württemberg Cooperative State University i Lörrach 26

27 ch.1: Vision (1) Cooperative Traffic Benefits Safety: (USA) 42,000 people die each year in traffic crashes, 2.7 million people injured. Significant share for pedestrian and other VRUs casualties fatalities Germany 2003 total pedestrian total pedestrians total ,60% ,30% city ,80% ,10% total rural ,50% ,70% highway ,30% ,90% Mobility: (USA) congestion based on traffic delays, wasted time and fuel cost taxpayers approximately $78 billion a year Environment: (USA) transportation accounts for roughly 28% of greenhouse gas emissions Infrastructure: (USA) required investments are estimated to between $225 and $340b annually Statistisches Bundesamt, Fachserie 8, Reihe 7, 2003, UJ 39(1) Source: S. Belcher, ITS America: Looking to the Future of ITS, Transpo2008, Orlando (FL) 27

28 ch.2: Basic Discussion Applications Examples Forwarding industry detection of objects i.e. random storage Farming tracing of animals Safety Detection of avalanche victims Military IFF - Identification i friend or foe Aerospace Secondary radar Traffic esafety 28

29 ch.2: Basic Discussion Cooperative Sensors use cases GIDAS (German In-Depth Accident Study)-Database: In 40%of all fatal pedestrian accidents the pedestrian was hidden to the driver until shortly before the collision happened. In 70%of all accidents with pedestrians, cooperative sensors could have supported the driver cited from: R. Rasshofer et.al. Pedestrian Protection Systems using Cooperative Sensor Technology, AMAA 07, pp , Berlin. peak of pedestrian fatalities in the time of o clock dawn SAFETYNET project, cited after: A. Mousadakou, A. Guarise, D.2.1 Requirements and use cases, WATCH-OVER project 29

30 ch.2: Basic Discussion IR Image Sensors Quelle: Daimler AG 30