January 04' 2020

09:00am-10:00am Keynote
Speaker: Prof. Shuvra S. Bhattacharyya
Title: Compact System-level Models for Design and Synthesis of Smart Sensing Systems
10.00am-10.30am Sensor-based Framework for Improved Air Conditioning Under Diverse Temperature Layout
Shadaab Siddiqie, Akhil Ralla, P. Krishna Reddy, Anirban Mondal, Aviruch Bhatia and Vishal Garg
(IIIT Hyderabad, Ashoka University and TERI School of Advanced Studies).
10:30am-11.00am A study of Approximation in a Collaborative Multi-Agent System
Hrishav Bakul Barua, Himadri Sekhar Paul, Chayan Sarkar and Ansuman Banerjee
(TCS Research and Innovation Labs and Indian Statistical Institute).
11:00am-11:30am Break
11:30am-12.00 noon Invited talk, Prof. Sanjay Purushotham, UMBC, USA.
12:00 noon-01:00pm Panel
Topic: The Future of Smart Sensing: Contact-full or Contact-less would be Prevalent
Panel Chair: Dr. Abhishek Mukherji, Accenture Inc., USA
Panelists:
Prof. Shuvra Bhattacharyya, UMD, USA.
Prof. Sanjay Purushotham, UMBC, USA.
Prof. Sandip Chakraborty, IIT Kgp, India.
Avik Ghose, TCS Research & Innovation, India.
Kishor Narang, Narnix TechnoLabs | Cron Systems, India.

Abstract:
Learning users' behavior in their homes and other environments is central to behavioral research. Social researchers are interested in how different family members spend their time at home, and how they interact. Further, healthcare professionals would like to understand patients' health conditions and caregiver-patient interaction. Current solutions to behavioral sensing rely mainly on self-reporting, i.e., having the subjects write diaries or answer questionnaires. These techniques, however, are often prone to subjective biases and inaccuracies. Moreover, keeping a detailed diary or administering questionnaires incurs a significant overhead from the subject and is not sustainable in long-term studies. An alternative approach would use wearable sensors or smartphones for behavioral sensing. However, several people could feel encumbered by wearables and uncomfortable using them. Further, past studies have shown that wearable devices lead to adherence problems because people stop using the wearable sensor with time. Thus, more recently, contact-less wireless sensing [1,2] without attaching a device to the target has achieved promising progress in recent years. Contact-less sensing uses the radio (RF) signals that bounce off people's bodies to enable behavioral sensing, without diaries or wearables. However, as of today, contact-less sensing has limitations such as small sensing range, and limited scalability in environments with multiple individuals. The panelists would discuss opportunities and challenges of the future of smart-sensing with respect to contact-full and contact-less sensing.
1. Enabling Identification and Behavioral Sensing in Homes using Radio Reflections, Dina Katabi et al. CHI 2019.
2. Fast and Contact-Free Object Detection using a Depth Sensor, Salil Kanhere et al. IEEE PerCom 2016.