131 



"teleoperators" to be general purpose submersible work vehicles controlled 

 remotely by human operators and with video and/or other sensors, power and 

 propulsive actuators for mobility, with mechanical hands and arms for mani- 

 pulation and, possibly, a computer for a limited degree of control autonomy. 



The objectives of the study are as follows: 



-Survey and analyze undersea tasks appropriate to accomplishment by tele- 

 operators. 



-Analyze constraints in the undersea environment and technological con- 

 straints of submersible vehicles, communication and control systems which 

 most significantly mediate teleoperator control - parimarily the man- 

 machine aspects. 



-Investigate and define theories of operator control performance applicable 

 to remotely controlled systems. Develop taxonomic and mathematical models 

 of man-machine interactions in undersea teleoperation (inspection, vehicle 

 control, manipulation), particularly those pertaining to supervisory control 

 where man controls computer on slow time scale while computer controls 

 teleoperator on fast time scale. 



-Recommend specific laboratory simulation experiments with human subjects 

 and software developments to explore and demonstrate various supervisory 

 control modes, and measure teleoperator performance. 



-Perform some of the above experiments and apply some of the above models. 

 (This is planned for follow-on phases of the present contract.) 



The following is an abstract of their work to date as presented in the July 

 1978 report: 



"This is a review of factors pertaining to man-machine interaction in 

 remote control of undersea vehicles, especially their manipulators 

 and sensors. Emphasis is placed on human operator control of such 

 teleoperator systems as a function of degree of automation sensor- 

 control integration and task demands for underwater search, object 

 recovery and manipulation. Models of operator-computer performance 

 are considered, particularly with respect to human supervisory control 

 of semi-autonomous systems. 



Sections of the report discuss: teleoperated submersible vehicles or 

 work platforms; undersea tasks and how they can be analyzed; relative 

 roles of human and computer or other control elements; control hardware 



(sensors, communication, propulsion, manipulation, control station) 

 and how it affects the human controller; control software for computer- 

 aided manipulation, including a review of various languages and algorithms 

 presently available; human operator performance in manipulator control 



(a review of what we now know) ; present and prospective theoretical 

 models of supervisory control; and finally, the needs for research 

 in this area. " 



As of late 1978 a laboratory model of a manipulative system has been fabricated 

 which is capable of measuring human input (such as the unscrewing of a nut 

 and its subsequent replacement on the bolt) and then duplicating this action 

 without the need for human intervention. 



