Infrastructure Sensor-based Automated Driving: Development and Demonstration

In this proposal, we intend to answer how infrastructure sensors can be used for autonomous driving. Using infrastructure sensors make automated driving safer, more simplified, and cost effective especially for multiple autonomous vehicles operating in known environments such as large residential/commercial complexes and resorts. Infrastructure sensors replace the main onboard vehicle perception sensors with infrastructure sensors mounted on the side of the road, for example on light posts. The information sent by all the sensor nodes to the cloud is used to localize all the autonomous and non-autonomous vehicles, pedestrians, and all other objects on the road for path planning and path tracking of any number of autonomous vehicles on the road. To enable infrastructure sensor-based automated driving, we will extend our navigation, object detection/classification, and motion planning/control modules from vehicle-installed sensors to infrastructure-installed sensors.

Faculty Supervisor:

Amir Khajepour;Ehsan Hashemi

Student:

Mobin Khamooshi;Ehsan Mohammadbagher;Neel Pratik Bhatt;Chao Yu;Amir Fathazam;Ruihe Zhang;Xin Xia

Partner:

Canadian Urban Transit Research and Innovation Consortium

Discipline:

Engineering - mechanical

Sector:

University:

University of Waterloo

Program: