Internet of Things (IoT)

Author: Thiago Teixeira (INRIA - France)

Date: 22/08/2011 (Mon)

Time: 10h

Location: IME-USP - Rua do Matão 1010 - auditório Jacy Monteiro

Abstract:

The Internet of Things (IoT) can be described as the extension of the Internet with sensors and actuators, with the aim to permit one to browse the real world with the ease and flexibility that are now staples of the virtual world. But in order to densely cover the real world, the number of such devices will certainly exceed the billions of units, and as a consequence the resulting flood of information within the Internet will make complex physical inferences as commonplace as a database query.

In this talk I discuss some of the greatest challenges related to the mixed real/virtual nature of the future Internet; Some have equivalents in the existing Internet, while others, concerning the physical aspect of the IoT, stem from fields such as robotics, estimation theory, and machine learning. I, then, propose directions for solving these challenges through the extensive use of approximations and automated estimations.

Biography:

Thiago Teixeira earned his Bachelors degree in Electrical Engineering from Johns Hopkins University (JHU) in 2003, with a double major in Mathematics. During that time, he started working at the Sensory Communication and Microsystems Laboratory under the guidance of Prof. Andreas Andreou, and obtained a Masters degree in 2005 researching Address-Event camera networks and acoustic localization wireless sensor nodes. Dr. Teixeira, then, joined Prof. Andreas Savvides's Embedded Networks and Applications Lab (ENALAB) at Yale University, earning his Ph.D. in Electrical Engineering in 2010. For his dissertation, he researched multi-modality sensing solutions to localize, track, and identify people within assistive home and office environments. The sensor network developed in this research was deployed in a number of real homes for several months at a time, aiding elderly people in their daily lives through activity-recognition. Dr. Teixeira is now a Post-Doctoral Associate in the ARLES team at INRIA Paris-Rocquencourt, working under Dr. Valérie Issarny. In his current work he is designing an automated sensor fusion engine for the Internet of Things. The goal is to empower users and developers of the future Internet to perform inferences that are currently only achievable by field experts — but to do so in an unsupervised, fully-automated manner.