Vivek Maru
Chemnitz University of Technology
Micro and Nano Systems
2. Semester
Autonomous Driving
Seminar Automotive Sensors
Chair for Measurement and Sensor Technology
Univ.-Prof. Dr.-Ing. O. Kanoun
2
Outline
• Motiviation
• Sensors Distribution
• System Architecture
• Multi-Sensor collaboration
• Sensors setup
• Data extraction and Observation
• References
Seminar Automotive Sensors
Chair for Measurement and Sensor Technology
Univ.-Prof. Dr.-Ing. O. Kanoun
3
Seminar Automotive Sensors
Chair for Measurement and Sensor Technology
Univ.-Prof. Dr.-Ing. O. Kanoun
What is Autonomous Driving? Why?
• An Autonomous car : A vehicle which is capable enough to fulfill human-
transportation capabilities without human inputs.
• Why do we need it?
• Accroding to The Eno Center for Transportations,Washington D.C,
- 90% of road accidents are actually human errors
- If 10 % cars are autonomous,
- 2,11,000 fewer crashes , 1100 lives can be saved p.a.
- If 50% cars are autonomous,
- 1.88 million fewer crashes, 9600 lives can be saved p.a.
BMW Next100, Image :
http://www.bmweducatiom.co.uk.
Google‘s AV, Toyota Prius.Image :
http://www.spectrum.ieee.org
4
Seminar Automotive Sensors
Chair for Measurement and Sensor Technology
Univ.-Prof. Dr.-Ing. O. Kanoun
Sensors Distribution
• Sensors distribution, Image by http://www.engineering.com
5
Seminar Automotive Sensors
Chair for Measurement and Sensor Technology
Univ.-Prof. Dr.-Ing. O. Kanoun
System Architecture
System structure, Image :
http://www.makezine.com
• Sensors :
- Raw input data
• ECU (Processor) :
- Decision Making
- Cheking functionalities of all
control drives
- Sending signals to actuators
according situation
- Alerts to Driver using UI.
• Actuators :
- Steering control
-Throttle control
-Differential control
- Motion contorl
- Break control
6
Multi-Sensor Collaboration
Seminar Automotive Sensors
Chair for Measurement and Sensor Technology
Univ.-Prof. Dr.-Ing. O. Kanoun
• A self-driving car, to be deployed in real-world driving environments, must be
capable of reliably detecting and effectively tracking of nearby moving objects.
• Sensor fusion for object detection
developed by Carnegie Mallon
University.
• LIDAR – For edge detection
• RADAR – For point detection
• Vision sensor / Camera – For
vision target
6x LIDAR
6x RADAR
3x Vision sensors / Cameras
• Final simulation screenshot, Image : http://www.cmu.edu
7
Sensors setup
Seminar Automotive Sensors
Chair for Measurement and Sensor Technology
Univ.-Prof. Dr.-Ing. O. Kanoun
• A radar – LIDAR pair at different
heights – Maximize the reliability and
range of measurements.
• 200 m – Any side of the vehicle
• 60 m – Object detected by at least 2
sensors (i.e. radar and LIDAR or radar
and camera)
• Sensor setup, Image : http://www.cmu.edu
• Coverage of sensors and blind spots, Image : http://www.cmu.edu
8
Data extraction and Observation
Seminar Automotive Sensors
Chair for Measurement and Sensor Technology
Univ.-Prof. Dr.-Ing. O. Kanoun
• High definition
screenshots of scenes.
• 3-D point clouds.
• Line segments or
junction of lines (“L”)
shaped.
• 2-D position and velocity
of objects.
• Precise measurements
even after 200m.
• Data from multiple sensors of same type – considered as 1 module.
• Sequential sensor method applied
a) Detection using camera, b) Detection using LIDAR, c) Detection using radar, Images : ICRA 2014, sensor fusion
9
Data extraction and Observation
Seminar Automotive Sensors
Chair for Measurement and Sensor Technology
Univ.-Prof. Dr.-Ing. O. Kanoun
a) Urben environment, b) Detection by radar, Image : ICRA 2014, sensor fusion
c) Detection by LIDAR, d) Detection by Vision sensors, Image : ICRA 2014, sensor fusion
10
References
• Daniel Fagnant and Kara M. Kockelman "Preparing A Nation For Autonomous Vehicles:
Opportunities, Barriers and Policy Recommandation" , October 2013
• Hyunggi Cho, Young-Woo Seo, B.V.K. Vijaya Kumar, and Ragunathan (Raj) Rajkumar “A Multi-
Sensor Fusion System for Moving Object Detection and Tracking in Urban Driving
Environments”, June 2014
• C. Mertz et al. Moving object detection with laser scanners. Journal of Field Robotics, 30(1), :
17-43 , 2013.
• June 2016 – http://www.cmu.edu
• June 2016 – http://www. http://driving.stanford.edu
• May, 2016 - https://www.enotrans.org/etl-material/preparing-a-nation-for-autonomous-vehicles-
opportunities-barriers-and-policy-recommendations/
• May, 2016 - https://www.engineering.com/electronicsdesign
• May, 2016 - http://www.makezine.com
• May, 2016 - http://www.bmweducation.co.uk
• May, 2016 - http://www.autotrader.ca
Seminar Automotive Sensors
Chair for Measurement and Sensor Technology
Univ.-Prof. Dr.-Ing. O. Kanoun
11
Thank you !!
Seminar Automotive Sensors
Chair for Measurement and Sensor Technology
Univ.-Prof. Dr.-Ing. O. Kanoun

Autonomous Driving- TU Chemnitz

  • 1.
    Vivek Maru Chemnitz Universityof Technology Micro and Nano Systems 2. Semester Autonomous Driving Seminar Automotive Sensors Chair for Measurement and Sensor Technology Univ.-Prof. Dr.-Ing. O. Kanoun
  • 2.
    2 Outline • Motiviation • SensorsDistribution • System Architecture • Multi-Sensor collaboration • Sensors setup • Data extraction and Observation • References Seminar Automotive Sensors Chair for Measurement and Sensor Technology Univ.-Prof. Dr.-Ing. O. Kanoun
  • 3.
    3 Seminar Automotive Sensors Chairfor Measurement and Sensor Technology Univ.-Prof. Dr.-Ing. O. Kanoun What is Autonomous Driving? Why? • An Autonomous car : A vehicle which is capable enough to fulfill human- transportation capabilities without human inputs. • Why do we need it? • Accroding to The Eno Center for Transportations,Washington D.C, - 90% of road accidents are actually human errors - If 10 % cars are autonomous, - 2,11,000 fewer crashes , 1100 lives can be saved p.a. - If 50% cars are autonomous, - 1.88 million fewer crashes, 9600 lives can be saved p.a. BMW Next100, Image : http://www.bmweducatiom.co.uk. Google‘s AV, Toyota Prius.Image : http://www.spectrum.ieee.org
  • 4.
    4 Seminar Automotive Sensors Chairfor Measurement and Sensor Technology Univ.-Prof. Dr.-Ing. O. Kanoun Sensors Distribution • Sensors distribution, Image by http://www.engineering.com
  • 5.
    5 Seminar Automotive Sensors Chairfor Measurement and Sensor Technology Univ.-Prof. Dr.-Ing. O. Kanoun System Architecture System structure, Image : http://www.makezine.com • Sensors : - Raw input data • ECU (Processor) : - Decision Making - Cheking functionalities of all control drives - Sending signals to actuators according situation - Alerts to Driver using UI. • Actuators : - Steering control -Throttle control -Differential control - Motion contorl - Break control
  • 6.
    6 Multi-Sensor Collaboration Seminar AutomotiveSensors Chair for Measurement and Sensor Technology Univ.-Prof. Dr.-Ing. O. Kanoun • A self-driving car, to be deployed in real-world driving environments, must be capable of reliably detecting and effectively tracking of nearby moving objects. • Sensor fusion for object detection developed by Carnegie Mallon University. • LIDAR – For edge detection • RADAR – For point detection • Vision sensor / Camera – For vision target 6x LIDAR 6x RADAR 3x Vision sensors / Cameras • Final simulation screenshot, Image : http://www.cmu.edu
  • 7.
    7 Sensors setup Seminar AutomotiveSensors Chair for Measurement and Sensor Technology Univ.-Prof. Dr.-Ing. O. Kanoun • A radar – LIDAR pair at different heights – Maximize the reliability and range of measurements. • 200 m – Any side of the vehicle • 60 m – Object detected by at least 2 sensors (i.e. radar and LIDAR or radar and camera) • Sensor setup, Image : http://www.cmu.edu • Coverage of sensors and blind spots, Image : http://www.cmu.edu
  • 8.
    8 Data extraction andObservation Seminar Automotive Sensors Chair for Measurement and Sensor Technology Univ.-Prof. Dr.-Ing. O. Kanoun • High definition screenshots of scenes. • 3-D point clouds. • Line segments or junction of lines (“L”) shaped. • 2-D position and velocity of objects. • Precise measurements even after 200m. • Data from multiple sensors of same type – considered as 1 module. • Sequential sensor method applied a) Detection using camera, b) Detection using LIDAR, c) Detection using radar, Images : ICRA 2014, sensor fusion
  • 9.
    9 Data extraction andObservation Seminar Automotive Sensors Chair for Measurement and Sensor Technology Univ.-Prof. Dr.-Ing. O. Kanoun a) Urben environment, b) Detection by radar, Image : ICRA 2014, sensor fusion c) Detection by LIDAR, d) Detection by Vision sensors, Image : ICRA 2014, sensor fusion
  • 10.
    10 References • Daniel Fagnantand Kara M. Kockelman "Preparing A Nation For Autonomous Vehicles: Opportunities, Barriers and Policy Recommandation" , October 2013 • Hyunggi Cho, Young-Woo Seo, B.V.K. Vijaya Kumar, and Ragunathan (Raj) Rajkumar “A Multi- Sensor Fusion System for Moving Object Detection and Tracking in Urban Driving Environments”, June 2014 • C. Mertz et al. Moving object detection with laser scanners. Journal of Field Robotics, 30(1), : 17-43 , 2013. • June 2016 – http://www.cmu.edu • June 2016 – http://www. http://driving.stanford.edu • May, 2016 - https://www.enotrans.org/etl-material/preparing-a-nation-for-autonomous-vehicles- opportunities-barriers-and-policy-recommendations/ • May, 2016 - https://www.engineering.com/electronicsdesign • May, 2016 - http://www.makezine.com • May, 2016 - http://www.bmweducation.co.uk • May, 2016 - http://www.autotrader.ca Seminar Automotive Sensors Chair for Measurement and Sensor Technology Univ.-Prof. Dr.-Ing. O. Kanoun
  • 11.
    11 Thank you !! SeminarAutomotive Sensors Chair for Measurement and Sensor Technology Univ.-Prof. Dr.-Ing. O. Kanoun

Editor's Notes