Program :    Biomedical Research and Countermeasures	Ground Research
Element :    Behavior and Performance
PI-in-a-Box
Principal Investigator:
Laurence R. Young, Sc.D. Aeronautics and Astronautics Building 37-219 Massachusetts Institute of Technology 77 Massachusetts Avenue Cambridge, MA 02139	Phone: (617) 253-7759 Email: LRY@MIT.EDU Fax: (617) 258-8111 Congressional District: MA-8
Co-Investigator(s):
Natapoff, Alan Szolovitz, Peter Massachusetts Institute of Technology Massachusetts Institute of Technology
Monitoring Center: NSBRI	Solicitation: NSBRI
Initial Funding Date: 1998	Expiration: 2000
Students Funded Under Research: 1	Post-Doctoral Associates: 0
Required Hardware:
Task Description:
This project tested an expert systems ability to meet the decision aiding needs of astronauts conducting life science experiments. By encapsulating the heuristic reasoning of a real principal investigator into P.I. in a Box, an expert system in an on-board computer, a computer decision aid can be provided. This project tested three hypotheses, as follows. When compared to conventional crew training alone, use of an expert system will improve average time to: (1) detect the deterioration of signal quality; (2) correctly identify the source of artifactual data in a complex situation; and (3) complete a normal calibration and run of a physiological experiment. The experimental protocol includes training individuals in the Neurolab sleep experiment with randomly degraded electrophysiological signals to test whether computer aiding improves training and subsequent performance. The expert system rules build on the existing [PI] programs for troubleshooting and error detection developed for the Neurolab Sleep experiment. Half the subjects were tested with [PI] assistance only on Day 1; the other half only on Day 2. For all subjects, time to detect and identify randomly introduced artifacts and to complete a normal sleep monitoring calibration was measured with and without [PI].
Phase 2 (1999-2000) of the study demonstrated a beneficial effect of [PI] and training in reducing anomaly troubleshooting time. Questionnaires showed that most subjects preferred monitoring the [PI] indicator lights while monitoring waveforms, rather than monitoring the waveforms alone. On one hand, [PI] did not improve the reliability of detection, since subjects were not any more correct in their anomaly detection with [PI] than without it. On the other hand [PI] did even out performance by reducing the chance of an undiagnosed fault, and by helping subjects with different tasks based on their experience level. It was shown that [PI]'s indicator lights only needed to be 40% reliable for subjects to achieve optimum performance, which shows its flexibility. [PI] correctly detected the anomalous signal for up to 85% of the time. There was no difference in fault management performance between genders.
While artificial intelligence and expert systems may serve as powerful tools for assisting experiments in space, the technology incorporated into a system such as [PI] is also relevant to any experiment situation where the Principal Investigator is physically remote from those conducting the experiment. Expert systems, which capture heuristic aspects of human knowledge and human methods to respond appropriately to unusual situations, have a flexibility that is highly desirable in circumstances where an invariably predictable course of action/response does not exist. An example of such a scenario could be the performance of an experiment in a remote and difficult to reach arctic location when the human expert back at the lab is unavailable, lacking the latest information, or is not consulted by those conducting the experiment. [PI]s benefit can also be extended from helping untrained astronauts to helping untrained sleep patients or caregivers fix problems with instrumentation. [PI] as the home sleep monitoring software would detect anomalous signals and suggest ways the patient or caregiver might fix the problem. [PI] could be a cost-effective way of improving the reliability of the home sleep monitoring system.
FY00 Publications, Presentations, and Other Accomplishments:
Atamer, A. ''Principal Investigator-in-a-Box Thesis.'' Massachusetts Institute of Technology (December 1, 2000). Delaney, M. ''Ground-Based Study of an Expert System for Human Assistance on the STS95 Sleep and Respiration Experiment.'' Massachusetts Institute of Technology (December 1, 2000). Atamer, A., and Delaney, M. ''Effectiveness of Principal Investigator-in-a-Box as an Astronaut Advisor for a Sleep Experiment.'' SmartSystems 2000 Conference (September 6-8, 2000). Atamer, A., Delaney, M., and Young, L.R. ''Ground-Based Study and Evaluation of Principal Investigator-in-a-Box.'' National Space Biomedical Research Institute Annual Conference (January 10-13, 2000).

Disclaimer:

Please note that certain ASCII special characters in the text, particulary in formulas, may not appear correct due to format limitations.