Projects by Principal Investigator
(sort by project name)
(sort by topic)
(project archive)

Learning-Oriented Dialogue in Cognitive Tutors: Towards a Scalable Solution to Performance Orientation
This project seeks to increase kids’ engagement with math learning in connection with Cognitive Tutors using support for help seeking and collaborative problem solving.

Vincent Aleven

Albert Corbett

Carolyn RosÉ

ACT-R
The ACT-R project is focused on the development of a robust and general cognitive architecture and its application to the modeling of human interaction with complex dynamic simulated environments and the creation of synthetic human-like agents.

ALPS: Active Learning and Problem Solving tutor
The Active Learning and Problem Solving (ALPS) project involves constructing and evaluating educational technology that emulates human tutors by integrating a state-of-the art educational technology called Cognitive Tutors with a innovative interactive questioning environment called Synthetic Interviews to produce an active learning environment that rivals the effectiveness of human tutors.

PACT--Pittsburgh Advanced Cognitive Tutor Center
The PACT Center uses cognitive tutor technology to create an integrated classroom and computer lab curriculum that supports students' understanding of mathematical and real world concepts. Based on a computational model of thought, cognitive tutors can automatically generate the most sensible solutions to any given problem, follow students step-by-step as they work, and provide individualized feedback and advice.

End-User Programming of Context-Aware Systems
Context-aware systems adapt to users' context of use. We are investigating novel interaction techniques to support end-users in building their own context-aware applications. In particular, we are supporting end-users in specifying complex Boolean context logic and demonstrating context-aware behaviors.

Anind Dey

IMPACT: Improving and Motivation Physical Activity through ContexT
A large percentage of the US population is overweight or obese, and a leading cause of this is lack of physical activity. Early fieldwork has shown that users have difficulty finding opportunities for physical activity and understanding how much exercise they get. To address these issues, we are exploring the use of mobile phone and web technology to monitor and inform useres abut their physical activity and the context in which the activities occur.

Anind Dey

Jodi Forlizzi

Improving Driving Routes through Learning Driver Preferences
Most route recommendation systems ignore the context of the driver (e.g. time of day) and the preferences of the driver (e.g. like to avoid highways). Not surprisingly, if a system can take this information into account, it can produce more appropriate routing directions. We apply novel machine learning techniques to learn driver preferences from actual driving data and use these to produce driving routes and predict driving destinations.

Anind Dey

Drew Bagnell

Intelligiblity of Context-Aware Applications
Applications that behave proactively on a user's behalf, particularly those that react to implicit user context, need to be intelligible to end users, explaining what they are doing and why. This project aims to improve the usability of and trust in context-aware applications, by gaining an understanding of how mental models are formed about context-aware systems, and designing interaction techniques and programming tools that will help application designers make their systems intelligible.

Leveraging Human Knowledge to Improve Learning
We are interested in understanding what a machine learning system could ask a user to improve its performance and when it is appropriate to ask the user. Our goal is to improve learning while minimizing user costs. We are exploring representations of a system's internal state and determining when a system could use input to improve performance

Memory Support for People with Episodic Memory Impairment
People with episodic memory impairment (EMI) lack an awareness of the actions, events and experiences of their everyday lives. This often results in confusion, depression, impaired decision-making, and additional stress for a caregiver. A common cause of EMI is Alzheimer's Disease. In this project, we are examining the use of capture and access technologies to improve recollection of past experiences and to reduce caregiver burden.

Creating Peripheral Displays
People who engage regularly with technology interact with hundreds of visual, auditory, and multimodal displays each day. These displays, which have been described as calm technology or peripheral or ambient displays, move information from the periphery to the center of human attention and back. Our team is working to determine the best sets of design variable to use to reduce the time it takes to extract information from a display.

Five features for human-robot interaction
This project explores how gaze, speech and sound, small movements, gesture, and proximity to the user affects the way that we perceive and work with social robots. We are testing these design primitives on everything from simple wheeled robots to humanoid systems.

The Hug
The Hug, a soft, huggable robot that uses sensing technology and wireless telephony, was designed to provide social and emotional support for elders who live at a distance from their family members. It provides intimate communication through voice augmented with touch, warmth, lights, and sound.

MOVE: Maps Optimized for Vehicular Environments
MOVE (Maps Optimized for Vehicular Environments) is an in-car navigation system that provides assistance for drivers who are navigating an unfamiliar route. It overcomes the limitation of current systems, which often provide too much information for the driver to attend to at any given moment. MOVE optimizes map information to provide situationally appropriate navigation information to the driver.

The Product Ecology Theory
The product ecology is a way to understand and design for different social relationships and experiences that people develop with intelligent products.

Jodi Forlizzi

The SenseChair
The SenseChair provides comfort and support for elders who spend long periods of time seated in the same chair, restoring independence and dignity to the life of someone who is nearly housebound. It motivates sitters to periodically move from the chair to stay mobile and active, and can provide assistance ranging from ambient reminders to explicit warnings.

Complex Collaboration
This project uses social science methods to better understand the requirements for successful multidisciplinary collaborations, the fundamental processes associated with geographic and functional distance, and applications that could reduce geographic and functional distance and improve conditions for successful multidisciplinary collaborations.

Gestures Project
In face-to-face settings, people use pointing and other gestures to communicate quickly and efficiently with their conversational partners. The aims of this project are, first, to understand how people use gesture to coordinate their talk and actions in collaborative physical (3D) tasks, and second, to develop and test systems to allow remote collaborators to gesture within a shared visual space.

Aura: Distraction-free Ubiquitous Computing
The focus of the research is on automatic configuration of an environment such as a device-rich room at all levels (networks/ devices, software infrastructure, HCI) to meet a user's needs and preferences. Configuration at the hardware level means that the network, devices, and other resources will be set-up on demand to support specific set of tasks.

Coordination in Open Source Software Development
Open-source software runs the Internet and challenges commercial software in many areas. Despite this success, it is not clear how the open-source paradigm solves problems of coordinating technical work over distance. This research is examining the processes of coordination in open-source software development to build a more general theory of coordination.

inTouch: Awareness and Messaging for Mobile Groups
inTouch is a mobile messaging platform that provides group-based communities (such as families, research work groups, carpools, etc) a better way to coordinate tasks, maintain a sense of shared awareness, and to generally keep in touch with fellow group members. inTouch aims to better address communication breakdowns that typically occur in short-term planning and coordination.

Marmite: End User Programming for the Web
The goal of Marmite is to make it easy to create "mashups" that combine content from multiple web sites and web services. Marmite lets end-users extract content from web pages, process it in a data-flow manner, and direct the output to a variety of useful sinks, such as displaying on a map.

Supporting Trust Decisions
People are increasingly being asked to make trust decisions online: should I open this email? Should I click on this link? Should I enter in my information? The consequences of a wrong trust decision can be dramatic, in terms of viruses, identity theft, and spyware. The goal of our research is to provide better tools, training, and user interfaces to help people make better trust decisions.

Jason Hong
Alessandro Acquisti
Lorrie Cranor
Julie Downs
Norman Sadeh

User-Controllable Privacy and Security for Pervasive Computing
Currently, it is hard to manage privacy and security in mobile and pervasive computing environments, both for system administrators and end-users. We are combining new kinds of user interfaces, visualizations, machine learning, and dialog technologies to give people a stronger sense of control over and feedback about the pervasive computing applications they interact with.

Designing Interfaces to Support Human Attention
Our research attempts to understand how to best design interfaces to support the demands and limitations of human attention. Currently, we are working to understand typical contexts in which attentional demands differ. Our investigations have focused on large displays in the environment as well as smaller mobile displays such as cell phones and PDAs.

Automatically adapting interfaces to user needs
We are working to create graphical user interfaces (GUIs) that automatically detect a user’s needs in order to automatically adjust the interface to accommodate them. We are focusing our efforts on understanding and detecting what types of errors people with motor impairments make when interacting with a mouse. Our approach is application and operating system independent.

Scalable Cognitive Modeling through Compositional Reuse
Cognitive modeling can scale affordably, and be routinely and efficiently applied to large complex tasks, only if it becomes an exercise of composing new behaviors from existing reusable behavior components. This research provides psychologically-validated reusable behavior components, a psychologically-validated theory of composition, and a high-level modeling language that incorporates this theory.

Usability and Software Architecture
The usability analyses or user test data are in; the development team is poised to respond. When the usability problems are presented, someone around the table exclaims, "Oh, no, its too late to change THAT!" U&SA explores how attention to usability at the software architecture design phase avoids this predicament.

Quality of Life Technology (QoLT)
The Quality of Life Technology (QoLT) Center is a National Science Foundation Engineering Research Center (ERC) whose mission is to transform lives in a large and growing segment of the population - people with reduced functional capabilities due to aging or disability.

Project on People and Robots
Autonomous mobile robots in the future will assist people in tasks that are physically demanding, unsafe, unpleasant, and boring. Service robots will need to live and work in a social world.This multidisciplinary project's goal is to develop a principled understanding of how to design robots that will accomplish social and instrumental tasks.

Manipulation in a Virtual Haptic Environment Based on Magnetic Levitation
Magnetic levitation is a substantially new approach which promises a qualitative leap in improvement of haptic (sense of touch) computer interactions. A user interacts with the computer by using a rigid tool to interact with computed environments. We are testing the capabilities of this system to guide manipulation.

Assistment Project
The ASSISTMENT project provides students with individualized
instructional ASSISTance while performing assessMENT. The research goal is to investigate whether we can provide students, teachers, parents and
school administrators statistically reliable and actionable data on
student progress without wasting student and teacher time on practice
tests that distract from learning.

Pittsburgh Science of Learning Center
Pittsburgh Science of Learning Center (PSLC) supports research on principles of “robust learning”. The Learnlab facility is designed to dramatically increase the ease and speed with which learning researchers can create the rigorous, classroom-based experiments that pave the way to an understanding of robust learning.

Rapid Development of Cognitive Models and Tutors
Cognitive Tutors have been successful in raising students' math test scores in high-school and middle-school classrooms, but their development requires considerable time and expertise. We are developing a set of authoring tools to make modeling both easier for experts and possible for novices in cognitive science. The tools draw on ideas of programming by demonstration, structured editing, and others. Careful application of HCI methods is key.

Coordinating attention and communication
A call during dinner? While one party may value the conversation, the receiver thinks it is an interruption. This project examines the conditions under which communication is valuable or disruptive and the social, economic and technological interventions that can ease conflict between potential parties to a conversation.

Designing online communities
Despite extensive experience, designing online communities remains largely trial and error. This project is developing theory to predict contribution behavior in online communities and developing theory-based guidelines for designing these communities. The project brings together researchers from Carnegie Mellon, the University of Michigan and the University of Minnesota with expertise in social psychology, economics, and computer science.

HomeNet
HomeNet is a research program to understand Americans' use of the Internet outside of the workplace. Starting in 1995, we have carefully documented how families use online services, how they are integrating electronic communication and information services into their lives and the impact these services are having.

Social Ties in the Internet Age
Project description: Proximity generally increases the likelihood of personal and work relationships, and geographic mobility disrupts them. Is this true in the Internet age? This research examines how information and communication technologies, such as cellular phones and the Internet, change the initiation, maintenance, and dissolution of personal, work, and community ties for recent movers. This research also attempts to understand what factors influence psychological and social adjustment to the new location after a residential move.

Footprints Project
The footprints project aims to encourage individuals to reduce their energy consumption by leveraging massive online social networks to show motivational information. Past work has shown techniques such as public commitment and group motivation to help encourage change. This interdisciplinary project combines work in human decision making, back channel communication, low-cost sensing, education about climate change and collaborative filtering to understand and address the problems inherent in changing energy consumption behavior. An important goal of the project is to deploy widely, reach critical mass, and study these factors in the field.

Glanceability Experiments
Glanceable visuals enable quick and easy visual information uptake, thus enabling users to monitor secondary tasks while they multitask or divide attention. However, little is known about how to best design visual information for divided attention situations. We present two experiments to address this question, which differ from past work in three ways: (1) We study information uptake speed for peripheral displays in dual-task situations; (2) we examine a wide range of renditions (graphic objects or text) inspired by existing displays, differing in both visual complexity and the degree to which they convey common meanings; and (3) we investigate how recognizable renditions are together as a set, and how this changes with different set sizes. Our main contributions are best practices for the design and evaluation of glanceable visuals, intended to help designers create better peripheral displays to support multitasking.

Project LISTEN's Reading Tutor
An automated Reading Tutor displays stories on a computer screen, listens to children read aloud, adapts Carnegie Mellon's Sphinx-II speech recognizer to analyze the student's oral reading, and intervenes when the reader makes mistakes, gets stuck, clicks for help, or is likely to encounter difficulty. mputer screen, listens to children read aloud, adapts Carnegie Mellon's Sphinx-II speech recognizer to analyze the student's oral reading, and intervenes when the reader makes mistakes, gets stuck, clicks for help, or is likely to encounter difficulty.

Natural Programming
The goal of the Natural Programming project is to make computer programming more accessible to novice, professional and end-user programmers. We are investigating how people think about interactive behaviors. Using these investigations, we are designing interactive programming tools and languages that are easier to learn and less error-prone.

Pebbles
The Pebbles project is exploring how Personal Digital Assistants (PDAs), such as the Palm Handheld or a device running the Microsoft Windows CE or Pocket PC operating systems, can be used when they are communicating with a "regular" personal computer (PC), with other PDAs, and with computerized devices such as telephones, radios, microwave ovens and factory equipment.

Alice
We are developing a tool called Alice that allows novice programmers to author interactive 3D virtual worlds. By identifying the unnecessary challenges of learning to program (such as syntax) and removing these challenges, we hope to make the fundamentals of programming accessible to middle school children.

Expanding the 3D Interaction
Lexicon
Much of the research on 3D interaction, particularly for immersive virtual environments, focuses on emulating the real world. Emulating the real world is not necessarily bad, but in a virtual environment we can do better. Our focus is on creating new interaction techniques that take advantage of the unique properties of virtual environments.

Infocockpits
Infocockpits: computer interfaces that improve human memory. Humans are adept at remembering information based on its location relative to their body and on the place where it was learned. Our interfaces include multiple displays surrounding the user, engaging human memory for location, and multimodal displays, engaging human memory for place.

Techniques for Interactive Audience Participation
Systems enabling large audiences to interact offer numerous possibilities for entertainment, education, and team building. We are developing interaction techniques that allow the members of a large audience to participate, either cooperatively or competitively, in shared interactive entertainment experiences.

CSR: Online Courseware in Causal and Statistical Reasoning
The great bulk of currently available "online courses" are no more than inconvenient textbooks. Our project is an ongoing effort to make online courseware that takes advantage of what the computer does well - interactive simulations, virtual laboratories, and allowing instant access to vast arrays of information and data.

Adapting to disaster
Hurricane Katrina had a profound effect on New Orleans and its music scene. Katrina devastated performance venues, dispersed bands and their audiences, and destroyed instruments. This project is a two-year, longitudinal study of how musicians used technology, both immediately after the storm and later, to cope with exile, scattering of family, friends and band members, and threats to their livelihood.

Mentored Maintenance: Use of Dialog and Advanced IETMs for Performance Enhancement Systems
This study builds on recent empirical work about user-centered design and effectiveness of high level IETMs (Interactive Electronic Technical Manuals) and the use of natural language dialog in tutorial systems. Natural language dialog and a second generation of high level IETM will be designed, prototyped, and field-tested for a well understood F/A-18 aircraft maintenance task.

ADEPT: Assessing Design Engineering Project Classes with Multi-Disciplinary Teams
This project seeks to improve the quality of learning in engineering design project courses by supporting the role of the instructor with an on-line monitoring/reporting system. The system predicts when groups need extra support through an automatic assessment of student discussions in an on-line groupware system called the Kiva.

Daniel P. Siewiorek

Susan Finger

Asim Smailagic

Carolyn Rose´

Context Aware Computing
Context is any information that can be used to characterize the situation of an entity (person, place, or physical or computational object). A computing system shold be aware of the person's context so that it can respond appropriately to user's cognitive and social state and anticipate their needs. Context Aware Computing will enable mobile computers to anticipate user needs, exploiting context information to significantly reduce demands on human attention.

DiMA: Diabetes Management Assistant Wearable Computer
DiMA is a wearable/handheld computer system with wireless communication and accessories, designed to help diabetic patients and their doctors to better manage the disease. We designed and implemented a prototype targeting diet, exercise, and medical monitoring. DiMA connects to the Internet; a web server is set up to allow patients to upload information to their personal web pages, where data mining can identify trends.

GM/CMU Project: Driver-Vehicle Interface
The goal is to improve security and driver-vehicle interfaces, by building a car that can analyze the driver's intention and watch the driver's physical and mental status for any impairments or information overload. The application combines a smart car environment and monitoring of driver state, with a wide range of input-output modalities.

Wearable Computers
Wearable Computers seek to merge the user's information space with their work space. Wearable research at Carnegie Mellon proceeds on several fronts. The Wearable Group develops new functional systems at a rate of about one design per year. Several of these systems require system or application software, some of which is distributed to the community.

Building Virtual Worlds
Building Virtual Worlds is not only taught to encourage working with other disciplines, but also is an experience with tools and process. Students use the best commercial 3D modeling and painting tools in conjunction with the Alice research project to push the boundaries and quality of tools available for artists.

Sonic Flashlight
Our objective is to permit in situ visualization of ultrasound images so that direct hand-eye coordination can be employed during invasive procedures. A method is presented that merges the visual outer surface of a patient with a simultaneous ultrasound scan of the patient's interior. The method combines a flat-panel monitor with a half-silvered mirror such that the image on the monitor is reflected precisely at the proper location within the patient. The ultrasound image is superimposed in real time on the patient merging with the operator's hands and any invasive tools in the field of view. Instead of looking away from the patient at an ultrasound monitor, the operator sees through skin and underlying tissue as if it were translucent.

CareMedia
CareMedia use behavioral capture, storage, and access techniques to create, maintain, and present a patient's behavioral log to care-givers. Our current focus is using video and audio capture to track disruptive vocalization in Alzheimer's patients in a dementia facility.

Pathway: Physics Teaching Web Advisory
The Physics Teaching Web Advisory (Pathway) is creating a proof-of-concept demonstration of a new type of digital library for physics teaching. It brings together several long-standing research projects in digital video libraries, advanced distance learning technologies, and collaboration technologies, and nationally known experts in physics pedagogy and high quality content.

Informedia Digital Video Library
Informedia is pioneering new approaches for automated video and audio indexing, navigation, visualization, summarization, search, and retrieval and embedding them in systems for use in education, health care, defense intelligence and understanding of human activity.

interACT
The international center for Advanced Communication Technologies is an international technology research lab. interACT is pioneering research in technologies that facilitate the human experience, human mutual understanding, and communication worldwide. Examples of such technologies are translation, speech, language, vision technologies, multimodal and cross-modal perceptual interfaces, smart rooms, and pervasive computing.