Platform Pittsburgh: Improving City Quality of Life Through City Scale Computing
As Pittsburgh becomes an increasingly instrumented city, this project seeks to harness the power of image sensors. The goal is to lay the groundwork to make it possible for researchers, businesses, local government and citizens to utilize this dramatic increase in data to develop applications that improve quality of life. The team has a set of goals to design and deploy a “living laboratory” for urban video analytics. As a first step, the team has developed an initial prototype of Platform Pittsburgh, which consists of multiple cameras and edge computers.
(Updated on September 7, 2018)
1. CMU researchers are working with the Department of Mobility and Infrastructure to deploy a prototype at a busy intersection in the City of Pittsburgh.
2. Platform created for capturing and processing high-resolution urban video streams that facilitates image processing/computer vision research at scale
3. Platform created for testing video analytics-based applications of interest to the City of Pittsburgh and to smart cities researchers. Support pilot studies to inform future at-scale deployments of video infrastructure in Pittsburgh and build local expertise on capabilities and value of urban video analytics.
4. Test bed developed for addressing privacy, data-management and policy challenges associated with dense, high-resolution urban video capture.
5. Establish guidelines/policies for data access and to enable authorized persons and organizations to access this dataset via a variety of programming APIs for future at-scale deployments in Pittsburgh and in other U.S. cities.
6. Developing methods for traffic analytics such as vehicle and pedestrian detection, 3D reconstruction of vehicles, and vehicle exhaust and wheel detectors.
This research is funded in parts by The Heinz Endowments, Intel, UTC-TSET, NSF, and Metro21.
Srinivasa Narasimhan, Professor, Robotics Institute
Robert Tamburo, Senior Project Scientist at the Robotics Institute