Fine grained "automatic vehicle classification" system development for accurately measuring passenger-freight interactions
This research proposal targets the design of vehicular road sensing networks used in the framework for Intelligent Transportation Systems (ITS). The PI will develop machine-learning models, optimized powersaving algorithms, communications protocols, and a low-power sensing platform to yield a novel and modular multi-node system for the purpose of “automatic vehicular detection and classification” (motorcycles, passenger cars, buses, trucks, etc.). The PI proposes to create smart highways by implanting wireless Micro-Electro-Mechanical System (MEMS) sensors, which will act like neurons to collect traffic data for vehicular movement. The proposed smart sensing and data interpretation system for smart roadways will be scalable, significantly cost-effective, maintain a small foot-print, and will be capable of detecting and classifying a vehicle in real-time. This research will focus on all levels of system design from architecture to computation to communication design.
P.I. Name & Address
Funding source: California Department of Transportation (Caltrans)
Funding amount: $87,703
Start date: 3/2/2017
End date: 6/30/2018