Menu

Menu controller icon

Walton researcher accepted for NGI Explorers Programme

Posted: 22-07-2021

Tags:

    Kanika Sharma, PhD Researcher

    Kanika Sharma, a PhD researcher at Walton Institute has been accepted for the NGI Explorers programme. The Next Generation Internet (NGI) is a European Commission initiative that aims to shape the development and evolution of the Internet into an Internet of Humans.

    The programme selects top European researchers and innovators to collaborate with universities in the United States for three or five month long expeditions to accelerate their projects. The program in its third call selected 23 researchers from all over Europe to work with US nodes to strengthen and articulate their ideas while building traction among a whole new network of partners and customers.

    Kanika’s proposal on ‘Using reinforcement learning to optimize the placement of distributed services on opportunistic clusters of moving vehicles’ has been funded for a collaboration with the Centre for Urban Informatics & Progress (CUIP) at the University of Tennessee at Chattanooga. This project is an extension of Kanika’s PhD research undertaken with her supervisors from Walton Institute, Senior Research Fellow, Dr Bernard Butler and Principle Investigator at CONNECT Dr Brendan Jennings. She was also supported by Walton Institute’s Head of Division (ENL) Dr Deirdre Kilbane and Postdoctoral Researcher Dr Dixon Vimalajeewa during the proposal writing and interview phase.

    The project is aligned with one of NGI’s five focus areas, Cloud/Edge Computing, based on technological trends that will thoroughly reshape the Internet over the next 10-15 years. This proposal aims to leverage closely moving vehicles in urban centres as an Internet of Vehicles (IoV) system by orchestrating the available resources to deploy data collection and video crowdsensing applications.

    Kanika explains the potential uses of this research which, “aims to deploy distributed applications like Convolutional Neural Networks (CNN)-based object detection for pedestrian safety, as well as to analyse traffic congestion and usage patterns in restaurants and service stations.”

    The proposal uses real vehicle density data to make intelligent decisions to place these services on moving vehicle clusters. This results in efficient data collection and processing, leveraging under-utilized embedded sensors and cameras on vehicles. The Fog computing approach also reduces the delay caused in uploading data to the cloud, which is critical for safety-related applications in a vehicular environment.