Project 1: EAGER: Creating a Community Infrastructure for Interoperable Emergency Connectivity
NSF Award ID : 1637371
Creating a Community Infrastructure for Interoperable Emergency Connectivity
First responders to disasters in the city need a complete picture of the community’s status
in order to accurately assess the condition of the inhabitants and organize available resources
to save lives, protect the environment and prevent further damage in the community. In localized emergency events, this information is collected through calls to the public safety answering point (9-1-1), through city employee radio communications (fire, law, public works, transportation, building inspectors), and through monitoring of public media and social media. However, under disaster conditions, these normal community services will be interrupted, including cell phones, internet connectivity and power. A disaster area might lose its connection to both local and regional emergency response assets, and existing rescue information networks may stop functioning resulting in efficiency degradation of rescue operations. In these circumstances novel systems must be available to substitute for the lost connectivity, to allow residents to contact to the public safety answering point, and to allow the Emergency Operations Center to collect and aggregate critical information across sectors to ensure that lifesaving operations are conducted expeditiously.
This research proposes to develop a community infrastructure for interoperable emergency connectivity that can operate in austere conditions, provide its own power, and create linkages throughout the community and across jurisdictional boundaries. The whole system is deployed on a two-tier architecture. The first tier is the edge device which functions as the participating gateways, pre-deployed in local communities with multi-modal communication modules as well as an external long range radio. The second tier is the central cloud deployed in the Emergency Operations Center, and responsible for resilient quality testing, automatic validation, disaster assessment, resource allocation, and the automation of remote edge devices.
Project 2: Smart Illegal Dumping Monitoring Service and Infrastructure in San Jose
In the streets of the City of San Jose, illegal dumping of trash, the abandonment of vehicles and illegal parking are big problems. This project aims to provide the City of San Jose with an efficient way to monitor illegal waste dumping, detect abandoned or illegally parked vehicles, and support environmental data collection using smart cameras and sensors that communicate through low-bandwidth city network infrastructure. We will design and develop prototypes of smart imaging, sensing and networking modules that contain cameras, embedded processors, and environment sensors and have wireless connectivity in the license-free UHF band, to relay data to a server in San Jose City data center using existing wireless mesh network nodes. This process will be automated and will result in a reduction of time needed to find and remove illegally dumped waste. This will eliminate odors and contaminants from the streets thereby protecting the environment and the health of citizens. The type of illegally dumped wastes varies from hot spots to districts, which can be construction trash, furniture, or household garbage. In order to recognize various forms of wastes, we will design a neural network model by feeding a deep neural network framework with thousands of training images. Various techniques to reduce communication overhead and processing energy will be developed.