Page:Advanced Automation for Space Missions.djvu/15

 automatic
 * 1) Reasoning or intelligence, including logical deductions, plausible inference, planning and plan execution, real world modeling, and diagnosis and repair in case of malfunction
 * 2) Man-machine interface, including teleoperator control, kinesthetic feedback during manipulation or locomotion, computer-enhanced sensor data processing, and supervision of autonomous systems.

1.3 Summer Study on Advanced Automation for Space Missions
Immediately following the conclusion of the Pajaro Dunes Symposium, the present summer study was convened on 23 June 1980 and completed its formal work (roughly 10,000 man-hours) on 29 August 1980. During the first two weeks of the study the group was introduced to the status of work in artificial intelligence by a series of lectures given by scientists from SRI International. A number of NASA program engineers participating in the study reviewed agency interests in relevant mission areas.

Study members then focused their work by selecting four space missions which appeared to have great potential for the use of machine intelligence and high relevance to future NASA program goals. There was no assumption that these specific missions would ever be carried out. The four teams and the missions they chose to examine were:

(a) Terrestrial applications - an intelligent Earth-sensing information system

(b) Space exploration - Titan demonstration of a general-purpose exploratory system

(c) Nonterrestrial utilization of materials - automated space manufacturing facility

(d) Replicating Systems - self-replicating lunar factory and demonstration.

The teams spent the major part of the summer elaborating their missions (summarized below), with particular emphasis on the special role that machine intelligence and robotics technology would play in these missions.

The study has produced three significant outputs, outlined briefly in the remainder of this chapter, as follows: Mission Scenarios, Advanced Automation Technology Assessment, and an Epilogue.

1.3.1 Mission Scenarios
Over the last few years literally hundreds of mission opportunities beyond the 10-year time frame have been developed by the NASA Office of Aeronautics and Space Technology and assembled into a comprehensive Space Systems Technology Model (OAST, 1980). To reduce the problem of automation technology assessment to manageable proportions, the summer study group formed four mission teams that could select single missions for concentrated attention in order to illustrate fully the potential of advanced automation. The task divisions among the teams guaranteed that all major classes of possible future NASA missions were considered, including public service, space utilization, and interplanetary exploration. A fifth group, the Space Facilities and Operations Teams consisted largely of NASA and industry personnel whose duty it was to ensure that all mission scenarios were technically feasible within the constraints of current or projected NASA launch- and ground-operations support capabilities.

(a) Terrestrial Applications Team. The Terrestrial Applications Team elected to examine a sophisticated, highly intelligent information processing and delivery system for data obtained from Earth-sensing satellites. Such a system can play an immediate and practical role in assisting people to manage local resources, and, in a broader sense, could provide continuous global monitoring that is useful in the management of the individual and collective activities of man. The mission scenario presented in chapter 2 includes basic systems descriptions and hardware requirements, a discussion of "world model" structures, and a suggested developmental timeline.

(b) Space Exploration Team. The Space Exploration Team developed the concept of a general-purpose, interstellar-capable, automated exploratory vehicle that can (1) operate in complex unknown environments with little or no a priori knowledge, (2) adapt system behavior by learning to enhance effectiveness and survivability, (3) independently formulate new scientific hypotheses by a process called abduction, (4) explore with a wide variety of sensory and effector-actuator systems, (5) coordinate distributed functions, and (6) exchange information with Earth via an entirely new form of man-machine interactive system. A demonstration mission to Titan was examined in some detail and is presented in chapter 3, including mission operational stages, hardware specifications, sensing and modeling functions, and machine intelligence and other advanced technology requirements.

(c) Nonterrestrial Utilization of Materials Team. The Nonterrestrial Utilization of Materials Team considered options for a permanent, growing, and highly automated space manufacturing capability based on the utilization of ever-increasing fractions of extraterrestrial materials. The major focus was the initiation and evolutionary growth of a general-purpose Space Manufacturing Facility (SMF) in low Earth orbit. The mission scenarios in chapter 4 include surveys of solar system resources and various manufacturing processes especially applicable to space, a description of several basic industrial "starting kits" capable, eventually, of evolving to complete independence of Earth materials resupply, and discussions of the rationales for and implications of such ventures.