Future Tools for the Home

Technology will revolutionize the tools that are used in the home of the near future. Embedded computing, sensing and actuation technologies—coupled with new infrastructure in the built environment itself—will combine to enable new opportunities to support the work of the home. Taking fullest advantage of these opportunities requires that we understand both the base technology on which these new tools will be built, as well as the domestic setting in which they will be used.

For example, robots are one category of home tools that people will likely increasingly purchase for their homes. However, to date very little is known about how to design robotic products that fit into the home, do meaningful work there, and are desirable to own. These are critical questions to answer if we are going to design usable and useful home tools for the future. Understanding how current home tools (including devices such as the iRobot Roomba vacuum cleaner) fit into the work of the home can shed light on how to design the next generation of domestic robots. Such understanding can also inform the creation of new low-level technologies that can be built into tools, or into the home infrastructure itself.

Current Projects

Energy Consumption Display

Researchers: Tae Jung Yun, Gregory Abowd
We investigated the impact of an in-home Energy Consumption Display (ECD), both stationary and portable versions, on household energy awareness and consumption. ECDs were deployed eight homes for three weeks each, providing half of the participants with a portable version and the others with a stationary one. Users reduced energy consumption by an average of 11% by identifying high-power devices in their home and by playfully setting conservation goals. Users who reported having low or moderate awareness of energy consumption at the outset of the study benefited most from the display. We found that the portable ECD better supported house-wide experimentation, but settled into stationary use after the initial survey of the home. We conclude by discussing the implications of these findings for the design of in-home ECDs.

Family Telepresence

Researchers: Lana Yarosh, Gregory Abowd
Non-residential parents and children in divorced families struggle to stay in touch. We are investigating the needs of these families to inform the design of technological intervention in this domain. The ShareTable is one such system. It combines videoconferencing and a shared tabletop surface to help children and adults who live apart stay in touch through natural activities like playing games, reading together, and doing homework.

Powerline Positioning (PLP)

NOTE: This research is transitioning into commercial product under the name Usenso.
Researchers: Erich Stuntebeck, Tom Robertson, Sidhant Gupta, Mayank Goel
Powerline Positioning is an indoor localization system capable of tracking objects, people and assets by leveraging the existing infrastructure in a building - power lines. A special module inject signals into the power line which is picked up by a tag (attached to subject being tracked) capable of calculating its position by fingerprinting these signals. The position information can then be used by any location aware application working on top of PLP.

Powerline Event Detection (PED)

Researchers: Gregory Abowd, Tom Robertson, Shwetak Patel, Sidhant Gupta, Mayank Goel, Yi Han

Activity sensing in the home has a variety of important applications, including healthcare, entertainment, home automation, energy monitoring and post-occupancy research studies. Many existing systems for detecting occupant activity require large numbers of sensors, invasive vision systems, or extensive installation procedures. PED uses a single plug-in sensor to detect a variety of electrical events throughout the home. This sensor detects the electrical noise on residential power lines created by the abrupt switching of electrical devices and the noise created by certain devices while in operation. Machine learning techniques are used to recognize electrically noisy events such as turning on or off a particular light switch, a television set, or an electric stove.

Viz-A-Vis

Researchers: Mario Romero, Gregory Abowd
Viz-A-Vis is a visualization tool for understanding human activity in natural environments over long periods of time. It is a sensing, interpreting and visualizing tool for analysis. Overhead cameras sense the human activity in natural environments. The data from the overhead cameras is interpreted by computing motion and blob-tracking. All levels on interpretation are visualized in a geographical information system where the floor plan of the inhabited space serves as the geography and time is stacked as layers on top of the plane. The goal of the tool is to contextualize video analysis in the space and time where target behaviors occur in order to facilitate rapid overview, filtering and zooming, and details on demand of large volumes of video data with potentially sparse target behaviors.

Home Infrastructure and Location Aware Projects

Capture Resistant Environment

Khai Truong, Shwetak Patel, Jay Summet, Gregory Abowd
The Capture Resistant Environment uses cameras and projectors to prevent unauthorized photography and video recording. The camera can detect the lens of a digital camera and the projected light can neutralize the camera, making any images or video recordings blurred and thus useless.
http://www.cc.gatech.edu/~summetj/cre

iCam

Shwetak Patel, Jun Rekimoto, Gregory Abowd

Effective integration of sensing and laser-assisted interaction have resulted in a handheld device, the iCam, which simultaneously calculates its own location as well as the location of another object in the environment. iCam demonstrates how location-aware, at-a-distance interaction simplifies certain location-aware activities.

Powerline Positioning

Shwetak Patel, Khai Truong, Erich Stuntebeck, Gregory Abowd
PowerLine Positioning (PLP) is an inexpensive technique that uses
fingerprinting of multiple tones transmitted along the powerline to achieve subroom-
level localization. We have compared PLP to other fingerprinting techniques and found that it compares favorably. More information on Powerline positioning

Studying the Proximity of Mobile Devices to Users

People: Shwetak Patel, Julie Kientz, Lana Yarosh, Gregory Abowd
Many Ubicomp and mobile applications being developed assume users alway have their phones nearby, but we suspect that this may not always be true. Thus, we are empirically studying the proximity of users to mobile devices to determine people's usage of mobile phones.
 

The Tableaux Machine at the Aware Home/Alien Presence

Mario Romero, Zach Pousman, Michael Mateas
Alien Presence is an AI-based interpretation of human activity with expressive, ambient, physical displays. The idea is to explore the subjective affordances of context aware ubiquitous devices. Do these devices simply help us juggle more tasks on our to-do lists, or can they help us to experience our own lives in new, deeper and more meaningful ways. We are specifically working on a system called the Tableau Machine. It is similar to Michael’s Office Plant #1, but instead of recognizing patterns in the inbox of the user email using NLP, it recognizes the “mood” of a household (the Aware Home) with low level perception. It maps the perception to autonomously constructed physical, miniature, abstract tableaux.

The Beware Home

Thad Starner, Maribeth Gandy, Andy Quay, Blair MacIntyre, Gregory Abowd, Cory Kidd
In the Broadband Institute Residential Laboratory, we are exploring interaction techniques for a contextually aware home [1]. Here we describe several recent projects that were adapted to create a haunted house for demonstration to the International Symposium for Wearable Computers conference during the month of October. These projects included three augmented realities, a location system, and five methods of interacting with the home environment.

TrackSense

Shwetak Patel, Moritz Koehler, Jay Summet, Erich Stuntebeck, Gregory Abowd
While commercial solutions for precise indoor positioning exist, they are costly and require installation of additional infrastructure, which limits opportunities for widespread adoption. We have developed a self-contained solution to precise indoor positioning that requires no additional environmental infrastructure. Evaluation of our prototype indicates that such a system can deliver up to 4 cm accuracy with 3 cm precision in rooms up to five meters squared, as well as 2 degree accuracy and 1 degree precision on orientation.

Virtual Rear Projection

People: Jay Summet, Gregory Abowd
Creating interactive surfaces with rear projected properties using multiple front projectors, allowing flexible deployment of interactive wall sized displays. This research has led to the release of the GVU PROCAMS toolkit, designed to ease the construction of projector camera systems.
http://www.cc.gatech.edu/cpl/vrp/

Home Networking

Tangible UIs for networking

Much of the home networking infrastructure is currently invisible, meaning that householders often have few affordances for how to control their networks. What if instead we provided tangible interfaces to home networking? We are developing tangible interfaces to facilitate more intuitive and richer interactions by making the home network more direct and manipulable. This project is exploring ways to incorporate the tangible interface into various home network elements including network setup and management, security, service composition among multiple networked devices, and media sharing among homes.

Ambient router

People often hide their home networking equipment in closets or under furniture. Rather than having networking equipment something to hide away, what if we could rebuild networking equipment so that it is fun, engaging, and something that you would want to have on display in your house?

Secure home network device provisioning

Householders report that device setup is one of the most challenging aspects of home networking. The ICEbox project is focused on providing easier mechanisms for provisioning client devices at the data link layer (enabling secure access to the network link), the transport layer (IP address assignment), and application layer (file shares, printer configuration), as well as more robust forms of network discovery, mapping, and management.

Home network visualization

The netviz project is exploring interactive visual representations for understanding and exploring the home network. The design of these visual representations is based in our earlier ethnographic fieldwork on understanding the needs and models of home network users, and uses a zooming metaphor to provide progressive revelation of information about network structure and detail.

Collaborative remote troubleshooting

When you've had problems with your home network, have you ever called a family member or friend for help? We're investigating ways to make it easier for people in different locations to communicate about their home networking problems and work together to fix them.

Customer service call analysis

We are analyzing a large dataset of actual calls to a home networking customer support telephone line. From this data, we hope to understand real-life challenges people face when setting up and fixing their networks.

Empirical studies of information sharing practices

n order to design better tools for home networks, we need to understand what homeowners are doing with their networks, rather than simply relying on our own intuitions. This project aims to learn about householder’s practices for information sharing on—and between—home networks, through a combination of home visits to actual users’ homes as well as structured interviews about their networks

End-user sketch analysis

People often have a difficult time describing infrastructure technologies such as networks. We are creating new research techniques to help people better communicate with researchers about the technologies they use in their lives.