homecurrent workfocus areas › human-machine interfaces

OVERVIEW:

human-machine interfaces   back to focus areas

Systems that help people interact with vast information, with new devices, and with each other


At PARC, we believe that we can help knowledge workers to use, manage, and share information more effectively by first improving our understanding of people and their work, of the structure of information, and of the capabilities of computers and networks.

To better understand people and their work, we use tools from ethnography, cognitive psychology, and human factors, through which we gain insights into how people handle information and communicate with each other at work and in other venues. This interdisciplinary blend of scientific approaches captures the wider context of human behavior as it pertains to interaction with technology.

To better understand the structure of information, we apply techniques from computational linguistics, image processing, document analysis, information retrieval, digital libraries, and knowledge representation. We combine these techniques with new software approaches that allow computers to act as intelligent assistants for individuals and teams, including activity detection, ubiquitous computing, web-based collaboration, and content-based networks.

With these approaches, we create tools that help individuals and teams work more productively with information as they find it, combine it, share it, and make sense of it.

PARC's methods are used to design technology with a thorough understanding of the meaningful activities for people in different settings — at work, at play, while traveling, and at home.

 

 

enabling technologies for industry applications

Conversation and interaction analysis to enhance work practices

Our ethnographic research has led to the creation of expert systems that support human decision-making, and software that recognizes conversational partners in large, multi-party audio streams. Our ethnographers are developing a new architecture to make robots interact more naturally with their environments.

Mobile device interaction

Unlike the attention that people tend to devote to using their desktop and laptop computers, we usually use mobile Internet devices — smartphones, netbooks, tablets — while in motion, often with one hand and while performing other tasks. Mobile devices do offer a wide range of interaction modalities (speech, video, touch screen, other sensors) that require deep knowledge of their relative tradeoffs. Our interaction research currently focuses on context-aware mobile services and high-performance touch-screen input, building upon a legacy of pioneering creation of mobile technologies, from the early days of the first handheld, touch-screen computer, the PARCtab.

Implicit interaction

Advances in sensor technologies and computer vision have opened a new interaction paradigm: implicit interaction. Without requiring us to directly drive the computer system to retrieve information, the system observes us and our environment, and infers the likely information that we need.

PARC research in implicit interaction combines insights from social, optical, and computer sciences. Examples include systems that use machine vision to detect a shopper's desires (for information) in a retail setting, and systems that predict a desktop user's information needs, based on information and people with which they recently interacted.