Thanuka Wickramarathne is a tenure-track Assistant Professor of Electrical & Computer Engineering at University of Massachusetts Lowell, Lowell, MA 01854 USA. Previously, he was a Research Assistant Professor in the Departments of Electrical Engineering and Computer Science & Engineering at the University of Notre Dame (Notre Dame, IN), DSP Engineering intern at Motorola Solutions (Plantation, FL) and Radio Network Planning Engineer at Celltell (Sri Lanka). He received his B.Sc. in Electronics and Telecommunication Engineering from University of Moratuwa (Sri Lanka) and both of his MS and PhD degrees in Electrical and Computer Engineering from University of Miami, Coral Gables, FL where he specialized in signal processing with an emphasis on multi-sensor data fusion. His current research interests are in the areas of multi-sensor data fusion, situational awareness, signal processing for big-data, and cybernetics with applications to defense, vehicular technology, smart infrastructure, and healthcare. He’s a Senior Member of IEEE, currently serves on the Board of Governors of IEEE Vehicular Technology Society and member of Administrative Committee (AdCom) of IEEE Sensors Council. Outside of professional work, he is an avid tennis player, amateur drummer, and enjoys traveling around the world with his wife.
I have worked on Dempster-Shafer theory and belief functions (in general) during my doctoral studies, with a primary focus on defense related soft and hard data fusion application under the support from US Department of Defense (I’ve attached a couple of papers where belief theoretic notions were utilized in advanced applications in machine-learning, control and optimization). Currently, my work falls under situational awareness, estimation and tracking, and trust in human-machine interactions with a focus on defense and transportation applications. Some of the challenges that my collaborators and I are having in the context of using the notions of Dempster-Shafer and belief functions in our communities (e.g., signal processing, vehicular technology, robotics, intelligent transportation) is the fact that except for one or two random cases, most reviewers are simply not receptive to the idea of using such modeling, either because they are hardcore Bayesians and/or unfamiliarity with the DS notions. This applies to Journal publications as well as proposals submitted to federal agencies. There are smaller groups in robotics and defense community who appears to be receptive, but in my experience, that’s a very small percentage.
I’ve been an active participant in several conferences, including International Conference in Information Fusion (FUSION), IEEE Vehicular Technology Society Conference (VTC), IEEE Intelligent Transportation Systems Society Conference (ITSC) and couple of other signal processing conferences (e.g., ICASSP, GloabalSIP) for the last 4–5yrs. I experience a similar pattern at these conferences. Therefore, I would envision a plan to create some activities, perhaps through my current affiliations with IEEE to approach the general fusion community and other data-driven communities to enlighten them of the advantages of belief functions and how they are related to the general probabilistic notions to improve acceptance and visibility of research in this domain. In a recent small journal article appeared in IEEE Sensors Letters special issue on Multi-Modal Data Fusion, I was trying to do exactly this (see Ref4). Another small article I had sent to FUSION conference in 2017 was also very highly received by the reviewer with glowing reviews (5/5 by 3 most of them), even though the actual contribution was minimal (Ref5).