DescriptionTraffic and congestion is one of the pressing problems facing modern society. Traffic congestion cost in the US is an estimated $166 billion and more than 3.3 billion gallons of fuel is wasted in stalled traffic.
A study performed by TrafficVision and the Clemson DICE Lab, shows how over a million hours of highway video can be processed in the Google Cloud Platform requiring over a million concurrent vCPUs of processing. This talk will discuss how the large scale traffic video analysis using Cloud HPC in the Google Cloud Platform can help agencies better understand congestion and incident hot spots within the camera coverage areas. This HPC process is also used for neural net auto-annotation allowing for AI algorithm improvements helping to provide better insights in the future.
We will also discuss the impact of real-time parallel video analytics on the transportation industry. How real time incident detection through video analytics is leveraged to reduce incident response time for emergency responders saving lives and reducing the duration of congestion periods. This capability also provides for roadway situational awareness across a broad transportation network.