ArticlePDF Available

Human and Computer Control of Undersea Teleoperators

Authors:

Abstract and Figures

The potential of supervisory controlled teleoperators for accomplishment of manipulation and sensory tasks in deep ocean environments is discussed. Teleoperators and supervisory control are defined, the current problems of human divers are reviewed, and some assertions are made about why supervisory control has potential use to replace and extend human diver capabilities. The relative roles of man and computer and the variables involved in man-computer interaction are next discussed. Finally, a detailed description of a supervisory controlled teleoperator system, SUPERMAN, is presented.
Content may be subject to copyright.
A preview of the PDF is not available
... some task T, is said to be non-autonomous from a j with respect to T; note that a i can be autonomous from a j in the context of other tasks. Sheridan and Verplank have proposed the 10-levels of autonomy scheme to describe the interaction of the involved agents based on the level of autonomy (Sheridan & Verplank, 1978). At the lowest level, an agent is merely a slave that does what is required from a human operator. ...
Article
Full-text available
In recent years, autonomous systems have become an important research area and application domain, with a significant impact on modern society. Such systems are characterized by different levels of autonomy and complex communication infrastructures that allow for collective decision-making strategies. There exist several publications that tackle ethical aspects in such systems, but mostly from the perspective of a single agent. In this paper we go one step further and discuss these ethical challenges from the perspective of an aggregate of autonomous systems capable of collective decision-making. In particular, in this paper, we propose the Caesar approach through which we model the collective ethical decision-making process of a group of actors—agents and humans, as well as define the building blocks for the agents participating in such a process, namely Caesar agents. Factors such as trust, security, safety, and privacy, which affect the degree to which a collective decision is ethical, are explicitly captured in Caesar . Finally, we argue that modeling the collective decision-making in Caesar provides support for accountability.
... High levels of human control AND high levels of automation are possible: The first idea is a fresh way to think of technology design, and breaks out of the traditional one-dimensional view that led readers to believe that more automation meant less human control. This one-dimensional view was first described by Sheridan and Verplank (1978), even though Sheridan (1992Sheridan ( , 2000 continued to have strong views about the need for supervisory control. ...
... Shared control was first proposed by Sheridan and Verplank, and was defined as "a situation where both automation and humans work on the same task at the same time" [25]. It is generally believed that intelligent control systems are good at the execution of fine tasks, while humans are skilled in the judgment and upper-level planning. ...
Article
Full-text available
This paper studies the lateral control problem for intelligent vehicles based on the concept of shared control. Considering the participation of drivers in the control loop, a shared control-based lateral controller is designed, which consists of two differed controllers: one is an LQR-based autonomous driving controller and the other is a driver’s intention-based fuzzy controller. For the vehicle dynamic model with two-degrees of freedom, an autonomous driving controller based on LQR and a driver’s intention-based fuzzy controller are designed. Then, the lateral controller based on shared control is constructed, which integrates the aforementioned two controllers. Finally, the co-simulation by MATLAB/Simulink and Carsim is presented. Furthermore, simulation results show that the designed lateral controller can track the desired path with better performance than the LQR-based autonomous driving controller.
... The first levels of automation were set by Sheridan in [9]. Here, 10 different levels described the amount of responsibility for each decision maker. ...
... L'une des plus anciennes définitions que l'on puisse trouver dans la littérature est celle proposée par Sheridan et Verplank [29] et reprise dans différents travaux [30], [31]. Les auteurs ont identifié six fonctions qu'un opérateur humain ou un ordinateur peuvent exécuter durant une opération : Requests (demande), Gets (récupère), Selects (sélectionne), Approves (approuve), Starts (commence) et Tells (informe). ...
Thesis
Au cours des dernières années, l’automatisation a été adoptée davantage par les fabricants. Ainsi, pour atteindre une performance industrielle et faire face à la concurrence, l’une des solutions pour chaque fabricant est de rationaliser l’automatisation. En effet, lorsqu’elle est conçue de manière appropriée, elle permet de réduire considérablement les coûts et résoudre de multiples problèmes liés à la qualité, la répétabilité ou encore l’ergonomie. Toutes ces données confirment qu’il est nécessaire de déterminer le niveau d’automatisation convenable pour chaque fabricant. Ceci permet d’atteindre les objectifs de l’industriel tout en évitant des investissements dans des solutions inappropriées.Ainsi, le travail de recherche a pour but de proposer une méthodologie d’aide à la décision pour améliorer le niveau d’automatisation des chaînes de montage automobile. Pour répondre à ce besoin, une analyse des travaux de recherche existants a été réalisée. Il en résulte qu’aucune des approches trouvées dans la littérature ne répond parfaitement à la problématique de recherche. D’autant plus qu’aucune des méthodes ne traite l’aspect du regroupement d’opérations et de l’équilibrage de la ligne. Ainsi, il a été essentiel de présenter une nouvelle méthodologie répondant aux exigences industrielles. La méthodologie développée comporte six activités qui permettent d’analyser l’ensemble des données d’entrée liées aux gammes de production, à l’architecture de l’usine ou encore à la conception des produits. Moyennant plusieurs critères de regroupements définis, la méthodologie permet d’étudier les contraintes de précédence et de proposer ainsi des regroupements d’opérations. Une fois, les regroupements d’opérations effectués, plusieurs scénarios d’automatisation sont proposés. Ainsi, afin de déterminer celui qui respecte le plus la vision de l’industriel, plusieurs critères de sélection ont été définis et quantifiés. L’applicabilité de la méthodologie a été vérifiée sur trois études de cas industriel en coopération avec notre partenaire.
Chapter
Shneiderman’s Human-Centered Artificial Intelligence (HCAI) framework suggests that high human control of automation is necessary to create reliable, safe, and trustworthy systems. The HCAI framework demonstrates that there is no need to sacrifice human control when incorporating higher levels of automation. We propose that Shneiderman’s two-dimensional framework is static and unable to incorporate contextual factors such as the decision for a human-in-the-loop system, cognitive limitations of the user, and user characteristics. The HCAI framework, while an essential foundation, ought to reflect the flexibility of AI systems, while meeting individual differences and situational requirements.KeywordsHCAIArtificial IntelligenceAutomation
Article
As autonomous systems become more complicated, humans may have difficulty deciphering autonomy-generated solutions and increasingly perceive autonomy as a mysterious black box. The lack of transparency contributes to the lack of trust in autonomy and suboptimal team performance. In response to this concern, researchers have proposed various methods to enhance autonomy transparency and evaluated how enhanced transparency could affect the people’s trust and the human-autonomy team performance. However, the majority of prior studies measured trust at the end of the experiment and averaged behavioral and performance measures across all trials in an experiment, yet overlooked the temporal dynamics of those variables. We have little understanding of how autonomy transparency affects trust, dependence, and performance over time. The present study, therefore, aims to fill the gap and examine such temporal dynamics. We develop a game Treasure Hunter wherein a human uncovers a map for treasures with the help from an intelligent assistant. The intelligent assistant recommends where the human should go next. The rationale behind each recommendation could be conveyed in a display that explicitly lists the option space (i.e., all the possible actions) and the reason why a particular action is the most appropriate in a given context. Results from a human-in-the-loop experiment with 28 participants indicate that by conveying the intelligent assistant’s decision-making rationale via the display, participants’ trust increases significantly and becomes more calibrated over time. Using the display also leads to a higher acceptance of recommendations from the intelligent agent.
Article
This paper quantifies the performance of an adaptive haptic shared control (HSC) paradigm in transferring the control authority between humans and automation when both agents face a conflict. Here, we considered the reverse-intent conflict between the driver and the automation system. We invited 27 participants to drive a simulated vehicle through a course wherein for sixty percent of obstacles, the human driver is instructed to avoid the obstacles in the opposite direction as the automation system was programmed. We employed a model predictive controller to determine the automation’s optimal impedance modulation policy so that the differential torque on the steering wheel is minimized while the obstacle is safely avoided. We compared the performance of the dynamic adaptive HSC with two other adaptive haptic shared control paradigms: assistive HSC and active-safety HSC. The assistive HSC paradigm represents a case where the automation has relatively high confidence in the driver’s actions. On the other hand, the active-safety HSC paradigm represents a case where the automation has relatively low confidence in the driver’s action. In the dynamic adaptive HSC, the automation adaptively assigns different policies to modulate its impedance based on the human’s adopted impedance. When a human’s adopted impedance is high, the automation in the dynamic adaptive HSC paradigms acts more like the assistive paradigm, and when the human’s impedance is low, the automation in the adaptive paradigms acts more like the active-safety paradigm. Here, we used the human grip force as a proxy to estimate the human impedance on the steering wheel. We compared the performance of these three haptic shared control schemes by analyzing five metrics, including obstacle hits and metrics related to driving maneuvers around the obstacles that were avoided. Our statistical analysis indicated that the dynamic adaptive HSC paradigm supports the best overall team performance in resolving a conflict between the driver and automation system while keeping the vehicle safe.
Article
It is shown that it is possible to predict aspects of manipulator performance from its kinematic geometry which is determined by the number of links and joints, as well as their types and sequence. The effect of manipulator geometry on each of the following is discussed: displacement of the free end, coupling of position and orientation, number of ways to reach a given position, solvability, working spaces, and approach angles.
Thesis
Thesis (M.S.)--Massachusetts Institute of Technology. Dept. of Mechanical Engineering, 1967.
Chapter
The use of man’s adaptive characteristics as a controller has a long history of providing a cost-effective method of increasing system reliabiity, from well before the Wright brothers’ first flights to the present day aviation and space missions. The basic technology was a primary limitation in earlier times but is less of a constraint now, and consideration must be given to such items as operator motivation and job satisfaction as well as performance and reliability.
Article
For an RUV to function near a target, it is necessary to maneuver the vehicle and stabilize it at any given point in space with a high degree of accuracy. By maneuvering, the manipulator's grabs are directed to the desired position, and through stabilization, this position remains nonvarying. Here, the central point is the manipulator/RUV interface. In connection with this, the problem of advanced RUV manipulator development appears to be a broad-based one whose solution requires an integrated approach which considers the interfacing of all RUV systems and which takes global RUV missions into account. This paper describes a method of RUV control, developed at the Institute of Oceanography (USSR Academy of Sciences) and based on an information dynamic model (IDM), has considerably improved control-process efficiency.
Article
In master-slave manipulators, forces encountered by the remote hand are transmitted back to the operator. At very great distances there will be a transmission delay between an operator's movement and a resulting force. Investigation was made of the effect of long delays and differences in strategy on positioning time with force feedback alone. Positioning could be accomplished, but delay coupled with high loop gain creates serious instability. Experimental results suggest that alternative displays of the feedback force can overcome the stability problem.
Article
Three studies were conducted to evaluate operator performance with conventional and stereo display systems. The first two studies involved perceptual judgment tasks (a modified Howard-Dolman depth discrimination test and a test of stereopsis, using random dot Julesz patterns), while the third employed a perceptual-motor task requiring end-effector positioning and closure. This third task was designed to approximate the major components of undersea object recovery missions. In studies one and two, two methods of displaying stereo information (Fresnel, Field Sequential) were compared. The results of study one indicate that both of these display systems provide adequate information to enhance performance over that given with a conventional monocular display. Study two indicates that thresholds of stereo viewing are comparable using the Fresnel system and the Field Sequential system. The data indicate no difference in thresholds over that obtained directly without viewer aids. In study three a conventional display was compared with the Field Sequential system used in studies one and two. The results indicate that the use of stereo cues reduces both response latency and errors significantly. An analysis of performance changes over repeated testing sessions indicates significant improvement on both variables. These effects, however, are non-differential across display systems, and are probably related to the acquisition of manipulator-specific motorskill learning.