1997 Fiscal Year Final Research Report Summary
Integration of Information in Cooperative Human-Machine Systems Learning in Gray Zone
| Project/Area Number |
07680446
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
社会システム工学
|
|---|
| Research Institution | University of Electro-Communications |
|---|
Principal Investigator |
TANAKA Kenji University of Electro Communications, 大学院・情報システム学研究科, 助教授 (60197415)
|
|---|
| Project Period (FY) |
1995 – 1997
|
| Keywords | Man-machine system / Evidence theory / Safety monitoring system / Human error / Uncertainty / Group of sensors / Time-criticality / Safety design |
|---|
| Research Abstract |
Results are classified into five categories as below.
1. From Bays probabilistic model to Dempster-Shafer model : Our research proposed a Dempster-Shafer theoretical model in which objective data were used and the optimal combination method were proposed to incorporate subjective belief information with objective information. The formulation in DS model is simpler than it in the Bays model so that the DS model is useful to learn effect of subjective information which human adds at a gray zone.
2. Role of human in cooperative human-machine monitoring system : In the DS model, a human-machine cooperative monitoring system was compared with an automatic monitoring system. As a result, some conditions were analytically proposed under which incorporating human judgments into automatic monitoring was more effective than complete automatic judgments.
3. Influence analysis of time-criticality on human reliability : By focusing on how time-criticality influences human judgments at the gray zone, I
… More
analyzed how human errors depend upon decision of the monitoring type, that is, human-machine cooperative type or complete automatic type.
4. Positively utilizing a gray zone : Reducing a gray zone does not always mean reducing uncertainty. Our results show that, rather, correct judgment is possible by utilizing positively the gray zone. From the point of view, our model was applied to product safety design and to condition based-monitoring in maintenance management. Also, our study proposed a safety design in consideration of a gray zone as a new paradigm of design toward the 21st century.
5. Application to mobile sensors system and its reliability analysis : Monitoring systems consisted of mobile sensors were analyzed where each sensor outputted "gray zone" in addition to "safety" and "danger" In the system, when the mobile sensor outputs gray zone, it collects several sensors and judges the state after integration of information. This application showed that utilizing the gray zone is useful to obtain reliability. Less
|
Research Products
(26 results)
