Main Page

From Emerging Technologies in Location-Aware Computing

Class Description

MAS S61.2011: Emerging Technologies in Location-Aware Computing

J. Paradiso, with G. Dublon and B. Mayton (offices in E14-548)

Class meets on Thursdays from 10 AM to 1 PM in E14-525.

We've witnessed the revolutionary applications that GPS and outdoor location services have provided - for example, there's no longer any excuse for being lost, we're guided to nearby restaurants we might like, we know where our friends are, and an emergency response can be automatically guided to the scene of an accident. But what if we knew where everybody and everything was both indoors and outdoors to within centimeters? This promises to be another game-changer and open a wide world of promise, perhaps with some peril. This special topics class will explore exactly this issue, brainstorming and prototyping applications based upon precision location.

Emerging radio technologies, such as Ultra Wideband (UWB) transceivers, are on the verge of becoming very inexpensive and highly integrated, and accordingly will soon become widespread and deeply embedded into everyday devices. This will usher new capabilities that will bring the mobility revolution to another peak - prime among which is the ability to radiolocate essentially any device to within a few centimeters. Similarly, computer vision approaches an analogous threshold, where cameras and 3D sensors (such as the Kinnect) can be paired with other sensors and ubiquitous processing, similarly enabling fine-grained location as well as a wealth of information that can serve to anchor context engines that aim at deriving a user's state. As exciting as these new technologies are, the applications that they enable remain speculative. The purpose of this class is to engage the creativity of Media Lab students in conceiving and realizing projects that explore possibilities en! abled by this frontier.

We plan to provide students with emerging tools that are at the cutting edge of these technologies - candidates are a wearable or easily embeddable sensor node based on a UWB technology and another that is based on radio direction-finding that are being developed by the Media Lab's industrial partners. We will also provide embedded computers based around Intel's Atom processor that will be used in base stations and fixed node infrastructure (including support for cameras and 3D sensors), together with code developed by our group to enable these platforms to function together.

Our lectures will be centered on discussion of relevant literature in the field of localization, applied computer vision, proximity-based UI's, and ensuing privacy issues, together with sessions getting participants going with the tools we've introduced above. Several of the classes will involve guest lecturers working in this field, both from the perspective of technology and application. A reading list and other details will be posted on the class wiki.

Final Project Documentation