Presentation on theme: "DriveSense’14 NSF Workshop on Large-Scale Traffic and Driving Activity Data DriveSense’14, Oct 30-31, Norfolk, VA."— Presentation transcript:
DriveSense’14 NSF Workshop on Large-Scale Traffic and Driving Activity Data DriveSense’14, Oct 30-31, Norfolk, VA
Day II – Breakout Groups Group I – Sensing, Collection and Mining on Road Group II – Modeling, Simulation and Testbeds Group III – Mobility, Connectivity, and Emerging Services/Applications Goal: Come up with a set of recommendations/requirements about visions/directions to be forwarded to NSF
Breakout Group I – Sensing, Collection, & Mining Infrastructure Sensing vs. Crowd Sensing Incentives – What – How to evaluate Unified Tools – Data Processing/Analysis – Validation Large-Scale Sensing Data! – How Large should be large? – How to execute? Privacy/Security – Do we need to redefine privacy? – Privacy vs. Safety? Faulty Sources vs. Malicious Behavior? New Mining tools vs. existing one? – Is there any unique characteristics that defines driving data? Heterogeneity – Data type sources vs. Data types Role of giant companies (e.g., Google, Navtaq)
Breakout Group II – Modeling, Simulation & Testbeds What data/factors need to be collected behavior/human modeling How to capture driver compliance? Active Behavior vs. Passive Incentives – What (insurance!) – How to evaluate How Big the data should be for a credible model? Sensitive behavior data vs. Safety? Intra-vehicle data vs. Inter-vehicle data Do we need new simulation tools? – What? – How? Testbeds! – Do we need it? – How large? – How to? How Driverless vehicles would change game.
Breakout Group III – Mobility, Connectivity, Emerging Services/Applications Multi-Modal Transportation Correlation between different mobility sources (e.g., pedestrians vs. vehicles) DSRC (Short Range Communication) has been regulated – Impacts? Role of clustering Energy Constraints! Simulation vs. Testbed – What scale is good enough? Security Services! If you had access to traffic traces / driving data, how would you want to filter and access it? New enabled applications – What data is needed? What do we need to collect?
Day II – Breakout Groups Group I – Sensing, Collection and Mining on Road – Lead: Ahmed Helmy/Hesham Rakha/Nagui Rouphil/Moustafa Youssef Group II – Modeling, Simulation and Testbeds – Lead: Mario Gerla/Brian Park/Bryan Porter Group III – Mobility, Connectivity, and Services/Applications – Lead: Jinzhu Chen/Avinash Srinivasan/Michele Weigle
What does the community lack? What type of resources we should have Within a 5year/10year time frame, what is the ideal state of data sensing and collection? How does your idea or this ideal state in five years differ from today? What do you see as significant barriers to achieving this ideal state? What aspects of DriveSense research are good candidates for attention? What are the emerging areas? In your opinion, what is missing and needs to be provided to inspire high quality research? What notable advances in research in other fields (Computer Science, Electrical and Computer Engineering, Civil Engineering, etc.) might apply? – How can NSF encourage the transfer of innovations from other fields and the development of new innovations in DriveSense? How to enhance shared knowledge and resource utilization – What type resources NSF should encourage the development? – Repository? Things to Think About
What is the ideal relationship between academia, government and industry partners? How can partnerships or collaborations be fostered? – What projects NSF should invest in that foster building relationship? – How can NSF foster relationships between industry and academia to support innovation? What institutional, educational, and other barriers are preventing DriveSense to advance? As computer science, computer engineering, civil engineering, etc., grew what were the barriers to their growth? How can experiences gained in other fields be applied to DriveSense? How to enrich Education? How to continue this momentum? Summer workshops? – How can academia keep related content current at various education levels? – What role can NSF play in assisting academic institutions maintain this momentum? – What type of initiatives NSF should support to invest in research and curriculum development to close the gap between the definition and application of core concepts within and across disciplines? Collaboration between different funding organizations/agencies (NSF/DoT) Things to Think About
Summary Slides to be presented at the Wrap- up session Summary Report to be later integrated in full report. Outcome of Breakout Groups