Using Artificial Intelligence and Sensors to Quantify Mobility in Real-time

Date: 10/1/20
street in city with yellow bars
Principal Researchers: Dan Connors

Unit: Department of Electrical Engineering

Project Abstract:
While vast resources are being invested in the creation of autonomous vehicles, identical attention must be placed on making equal advances in smart, connected intelligent transportation infrastructure. SmartCity infrastructure enabled by artificial intelligence (AI) can perceive objects (e.g., vehicles, pedestrians, bikes, etc.) on the roadway and gather information on individual vehicles or composite state of traffic at a considerably finer level of granularity than present systems provide. The proposed seed grant will extend and promote the development of AI-based computer vision developed by Professor Connors and University of Colorado Denver’s Edge Computing Laboratory.

The core goal is to enable infrastructure-based computer visual perception and sensor fusion that quantify all mobility within transportation and urban areas. With an application to two areas of immediate practical interest the research will highlight using UC Denver Auraria campus as an open-research SmartCity environment: vehicle identification and classification, and smart intersection signaling. In both cases the project will utilize physically and visually realistic computer simulation to develop and evaluate deep learning neural network algorithms and follow with a pilot deployment of the algorithms for real-world validation with municipal partners. Overall, with the help of community partners, the seed funding will help extend existing work in artificial intelligence and computer vision further into the SmartCity domain. Seed support will impact the potential success of large-scale funding opportunities within National Science Foundation, Department of Energy, and Department of Transportation.