My name is Dewant Katare. I am working as an Early-Stage Researcher for APROPOS project since February 2021 at the Faculty of Technology Policy and Management, Delft University of Technology (Netherlands). I am from Gwalior, India. I have completed my bachelor’s degree in Electrical and Electronics Engineering in 2016 and master’s degree in Electrical and Computer Engineering in 2019 from Purdue University. While completing my master’s degree I have been fortunate to publish a few IEEE conference papers, and have developed a keen research interests in automated driving services, embedded systems, edge intelligence and vehicular communications.
The title of my PhD project is “Approximation at the Edge for Automated Driving”. The objective of this research is to develop an energy-aware software framework for edge devices using approximation technique and Edge AI processing pipeline. Automated driving is dependent on several services and applications which consumes significant amounts of energy and power through regular sensing, on-board computing and communication within the vehicular ecosystem. Some of these vehicle’s data-services and applications allow approximation opportunities by reducing the on-board energy consumption and by balancing the energy-accuracy tradeoff. In particular, the project will be concentrated on lossless data compression approaches, development and deployment of DNN models on resource constrained edge devices and software acceleration approaches. While developing models the focus is also given on exploring biasness through open access datasets.
I am very much fortunate to be part of APROPOS project and the PhD program. This gives me an opportunity to develop the skills and knowledge in applications for resource constrained devices and targeting them for achieving energy-efficiency by also exploring the socio-technical system challenges and perspectives.
