experimenter perspective towards a cloud
play

Experimenter Perspective: Towards a Cloud- Assisted Smart - PowerPoint PPT Presentation

Experimenter Perspective: Towards a Cloud- Assisted Smart Intersection Marco Gruteser WINLAB Rutgers University 2 ECE - Winlab HUMAN DRIVER SELF-DRIVING CARs ~100 million miles / fatality Billions of miles / fatality ECE - Winlab


  1. Experimenter Perspective: Towards a Cloud- Assisted Smart Intersection Marco Gruteser WINLAB Rutgers University

  2. 2 ECE - Winlab

  3. HUMAN DRIVER SELF-DRIVING CARs ≪ ~100 million miles / fatality Billions of miles / fatality ECE - Winlab

  4. Connectivity Enables Foresight Through Road Data Sharing Analysis of Months Near - Accidents / SW Updates Minutes Live HD Maps and Point Clouds Milliseconds Real-time Sharing ECE - Winlab

  5. BigRoad Mileage Goal Years 1,000,000,000 = 1 Mileage / Year Vehicle Num 100 × 100,000 10,000 100,000 100 ECE - Winlab

  6. Connectivity Enables Foresight Through Road Data Sharing Analysis of Months Near - Accidents / SW Updates Minutes Live HD Maps and Point Clouds Milliseconds Real-time Sharing ECE - Winlab

  7. Eliminating Blind Spots with Intersection Sensor Sharing ECE - Winlab

  8. Sharing Rich Sensor Data over V2V w/ R. Govindan (USC) ECE - Winlab

  9. Edge-assisted Intersection Model Edge computing resources Live Merged Point Cloud and Guidance ECE - Winlab

  10. 10 Tethered design in high-quality VR systems • Inconvenient Can we cut the • Limited mobility cord? • Hazards ECE - Winlab

  11. Contribution of our system • We design an untethered VR system that is able to achieve both low-latency and high-quality requirements over a wireless link. • Parallel Rendering and Streaming • Remote VSync Driven Rendering ECE - Winlab

  12. Scalable Vehicle-2-Vehicle Communication Simulator (CAMP VSC3, USDOT) ECE - Winlab 13

  13. Developing Model and Robust Protocols for Real Road Environments (CAMP VSC6, USDOT) ECE - Winlab

  14. ECE - Winlab

  15. COSMOS : Sensor Sharing Experiments Vehicular Sensor Automotive Research: Rich Sensor Sharing and Orchestration for Robust Sharing and Automated Driving automated driving • Experiment involves Collection and Months Analysis of Cloud analytics multiple mobile Near - nodes, high BW/low Accidents latency wireless Minutes access and multiple Edge computing levels of cloud resources Live Merged Point Cloud processing and Guidance • Real world traffic and network mmWave Milliseconds conditions access network • and direct Outcomes include communications evaluation of system performance and application demo ECE - Winlab

Recommend


More recommend