I-Corps: Advanced Truck Detection with Lidar Technology

点击翻译按钮获取中文摘要

The broader impact/commercial potential of this I-Corps project is to enable transportation agencies to gather critical freight movement data using passively collected and anonymous sensors. Anonymity is of key importance for the successful collection of transportation datasets, especially in the competitive freight industry. Traditional approaches such as image-based detection, cell phone tracking, or other visual monitoring (license plate tracking or logo recognition) can violate privacy considerations and hinder widespread freight data collection. Such data collection is necessary for travel demand modeling and forecasting as well as for infrastructure planning, operations, and maintenance for roadways, bridges, and freight terminals. The market for this advanced truck detection device includes public transportation agencies at the city, county, state, and national levels, traffic sensing device manufacturers, transportation consulting companies, and freight terminal operators. While current sensors may distinguish trucks from cars or trucks by axle configuration, there are no non-pavement intrusive technologies currently able to predict the body-type of the vehicle in enough detail to indicate freight carried. Agencies tasked with data collection often must rely on time-consuming periodic surveys to estimate where and what freight is moving on their highway system, making it difficult to produce timely project cost-benefit and resilience/impact analyses. Commercial applications can be extended to large distribution centers, mining areas, rail yards or other intermodal terminals and ports.This I-Corps project will further develop a system for low-cost, anonymous, and pavement-nonintrusive advanced truck detection by developing a side-fire (perpendicular to traffic flow) Lidar (Light Detection and Ranging)-based traffic detection and classification system. In side-fire configuration, Lidar sensors capture the profile of the truck (tractor and trailer/semi-trailer) which can be classified by body type with high-resolution while maintaining the anonymity of the driver, license/registration, and company. The innovation of this technology includes: 1) novel configuration and application of off-the-shelf Lidar technology for traffic detection, 2) coupling of technology with machine learning algorithms for feature detection, extraction, and classification with the aim of high-resolution truck classification, and 3) implementation of classification outputs in a data dashboard for real time and historical review. This novel truck detection solution using Lidar can enable a fundamental shift in how freight data is collected, especially by public transportation agencies.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.