Link to the Information Technology Laboratory Website Link to the Information Access Division Website Link to the NIST Website NIST, IAD Banner

Chapter 1. Introduction

The NIST Smart Space Project

The NIST Smart Space Laboratory has been created to offer assistance to industrial research and product development laboratories to face the numerous performance and interoperability challenges inherent in the Smart Work Spaces of the future. We believe that the shift to Pervasive Computing is already well under way and will have as much impact on industry and daily life as personal computing did. Our Modular Test Bed is designed to allow our industrial and academic laboratories to bring their technologies together for integration and performance testing needed.

Pervasive Computing refers to the trend to numerous, easily accessible computing devices connected to each other and to a ubiquitous network infrastructure. This will create new opportunities and challenges for IT companies as they place computers and sensors in devices, appliances, and equipment in buildings, homes, workplaces, and factories. Within a few years, embedded devices able to execute complex software applications, and use wireless communications will be the norm and their effective use requires distributed interfaces on numerous, small, and even invisible devices.

The second version of the data flow system focuses on advanced forms of human-computer-interaction, integrating wireless networks with dynamic service discovery, automatic device configuration, and sensor based perceptual interfaces. The objectives are to:

  • Identify security mechanisms needed to ensure privacy, integrity, and accessibility of implementations

  • Develop metrics, test methods, and standard reference data sets to pull the technology forward

  • Provide reference implementations to serve as models for possible commercial implementations

  • Establish an integrated multi-sensor perceptual interface test bed for smart environments

  • Deploy perceptual interface components from our industrial advisors to investigate system level performance and metrics issues

  • Provide a multi-sensor data recording environment for the production of standard test materials

This middleware makes it possible to integrate components that were not intentionally designed to work together, such as speaker identification, and speech recognition systems. Many constituent technologies are under separate development in industry, so the issues of interoperability and integration are paramount. This project will develop metrics and reference material based on real implementations for industrial use. Important technologies include sensor-based collaborative interfaces using:

  • Speech recognition

  • Speaker identification

  • Face recognition

  • Source localization/separation

  • Channel normalization

  • Immersive video and acoustic displays

  • Personal information appliances

  • Pervasive networking

  • Information storage and retrieval

  • Multimedia data types

  • Multiple Interconnected Spaces

Standardization and measurements are critical to foster integration, interoperability, and the development of emerging technologies but are often not addressed by individual companies for economic reasons. The lack of common software infrastructure, tools to create, manage, measure, test, and debug pervasive services, standards in key areas such as service discovery, APIs, wireless networks, automatic configuration, and ad hoc transactional security, all currently impede widespread adoption of pervasive computing. Also, measurements, and tests developed and performed in a public forum allow competing research systems to be compared and improved to build on the best features of previous iterations. It is a long-term research platform that will provide a sensor-rich collaborative working environment. Sensor technologies include microphone arrays, video camera arrays, smart badges, I.R. room scanners, and possibly person position sensors.

Created on 2008-06-18 by Antoine Fillinger - Last updated on 2008-11-23 by Antoine Fillinger