Roles for Agent Assistants in Field Science: Understanding Personal Projects and Collaboration

• Main Author: William J. Clancey
• Other Authors: The Pennsylvania State University CiteSeerX Archives
• Format: Text
• Language: English
• Subjects: A; Arctic
• Online Access: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.154.5277; http://www.ic.arc.nasa.gov/m/pub/547h/0547%20(Clancey).pdf

Abstract—A human-centered approach to computer systems design involves reframing analysis in terms of the people interacting with each other. The primary concern is not how people can interact with computers, but how shall we design work systems (facilities, tools, roles, and procedures) to help people pursue their personal projects, as they work independently and collaboratively? Two case studies provide empirical requirements. First, an analysis of astronaut interactions with CapCom on Earth during one traverse of Apollo 17 shows what kind of information was conveyed and what might be automated today. A variety of agent and robotic technologies are proposed that deal with recurrent problems in communication and coordination during the analyzed traverse. Second, an analysis of biologists and a geologist working at Haughton Crater in the High Canadian Arctic reveals how work interactions between people involve independent personal projects, sensitively coordinated for mutual benefit. In both cases, an agent or robotic system’s role would be to assist people, rather than collaborating, because today’s computer systems lack the identity and purpose that consciousness provides. Index Terms—Collaborative work, robots, model-based systems, field science, assistants, consciousness