Randall Shumaker

The importance of shared mental models and shared situation awareness for transforming robots from tools to teammates

Current ground robots are largely employed via tele-operation and provide their operators with useful tools to extend reach, improve sensing, and avoid dangers. To move from robots that are useful as tools to truly synergistic human-robot teaming, however, will require not only greater technical capabilities among robots, but also a better understanding of the ways in which the principles of teamwork can be applied from exclusively human teams to mixed teams of humans and robots. In this respect, a core characteristic that enables successful human teams to coordinate shared tasks is their ability to create, maintain, and act on a shared understanding of the world and the roles of the team and its members in it. The team performance literature clearly points towards two important cornerstones for shared understanding of team members: mental models and situation awareness. These constructs have been investigated as products of teams as well; amongst teams, they are shared mental models and shared situation awareness. Consequently, we are studying how these two constructs can be measured and instantiated in human-robot teams. In this paper, we report results from three related efforts that are investigating process and performance outcomes for human robot teams. Our investigations include: (a) how human mental models of tasks and teams change whether a teammate is human, a service animal, or an advanced automated system; (b) how computer modeling can lead to mental models being instantiated and used in robots; (c) how we can simulate the interactions between human and future robotic teammates on the basis of changes in shared mental models and situation assessment.

Anthropomorphism of Robotic Forms: A Response to Affordances?

Participants rated robotic forms on three scales: perceived aggression, intelligence, and animation. The robot bodies varied along five dimensions: Types of edges (beveled or squared), method of movement (wheels, legs, spider legs, or treads), number of movement generators (2 or 4), body position (upright or down), and presence of arms (present or absent). Across ratings, movement method and presence of arms were the strongest predictors of participant perceptions. Legs and arms, both human characteristics, were associated with more positive attributions. Minimal affective characteristics, as displayed by the body design, are important in user perceptions of use and ability.

Multi-agent task allocation: learning when to say no

This paper presents a communication-less multi-agent task allocation procedure that allows agents to use past experience to make non-greedy decisions about task assignments. Experimental results are given for problems where agents have varying capabilities, tasks have varying difficulties, and agents are ignorant of what tasks they will see in the future. These types of problems are difficult because the choice an agent makes in the present will affect the decisions it can make in the future. Current task-allocation procedures, especially the market-based ones, tend to side-step the issue by ignoring the future and assigning tasks to agents in a greedy way so that short-term goals are met. It is shown here that these short-sighted allocation procedures work well in situations where the ratio of task length to team size is small, but their performance decreases as this ratio increases. The adaptive method presented here is shown to perform well in a wide range of task-allocation problems, and because it requires no explicit communication, its computational costs are independent of team size.

Developing and Anthropomorphic Tendencies Scale

A 208-item scale was developed to measure self-reported anthropomorphic tendencies during interactions with various non-human entities. The potential targets of anthropomorphism included technology-laden machines such as computers, other objects such as backpacks, living things such as houseplants, and abstract entities such as a god or higher power. Scale items assessed the degree to which participants agreed with statements regarding the perceived attributes of the entities, speech directed toward the entities and the treatment of the entities. A factor analysis suggested that the scale measures four independent types of anthropomorphism: “extreme” anthropomorphic tendencies, anthropomorphism of a god or higher power, anthropomorphism of pets, and “negative” anthropomorphism. Further analyses indicated that anthropomorphic tendencies were self-reported when pertaining to pets and a god or higher power. However, participants tended not to self-report inappropriate “negativeâ” anthropomorphism toward computers, cars, microwaves, etc. These disparate findings appear to be due to social desirability of anthropomorphism.

Empirical Study on the Effects of Synthetic Social Structures on Teams of Autonomous Vehicles

The goal of this research is to explore the effects of social interactions between individual autonomous vehicles (AVs) in various problem scenarios. We take a look at one way to construct the social relationships and generate data from computer simulations to compare the behaviors of each. A difference can be noticed when synthetic social structures (SSS) are used to control the interactions between neighboring AVs. Our experiments show that SSSs can be used to improve team performance on a problem in which a team of AVs must maneuver through a narrow corridor to reach a goal

Size Does Matter: Automobile “Facial” Features Predict Consumer Attitudes

This study further examines the issue of whether perception of automobile “faces” can be predicted based on the way in which people process human faces. Specifically, this work examined whether consumer ratings are correlated with the relative size of features on the front end of an automobile. Participants rated twenty-three cars on a variety of consumer attributes. Greater distance between headlights, like distance between human eyes, predicted positive ratings on several consumer attitudes. Results are consistent with the idea that processes used for perception of faces are involved in the attributions made for artifacts displaying minimal features.

When Function Follows Form: Anthropomorphism of Artifact “Faces”

Participants rated machine “faces” which varied in terms of eye size, eye shape, distance between eyes, and relationship to background color (white on black or black on white). Ratings were made for aggression, friendliness, intelligence, trustworthiness, and degree of animation. In addition, reaction time was collected for all ratings. Large, round, and close-set eyes were perceived most negatively across ratings. Aggression ratings were predicted by simple variables, whereas trustworthiness ratings were predicted by interactions among variables. Some judgments of form require the assessment of specific features, whereas others rely on a “gestalt” assessment of many features simultaneously. Humans attribute personality characteristics to minimal features, suggesting that form of intelligent artifacts is important in predicting human interactions with that item.

Effects of Eye Structure and Color on Attributions for Intelligent Agents

Participants rated machine “faces” which varied in terms of facial feature shape, whether the eyes had a pupil or not, and seven eye colors (blue, green, orange, pink, red, purple, or yellow). Ratings were made for aggression, friendliness, intelligence, trustworthiness, and degree of animation. Faces with eyes that had a discernable pupil were rated most positively, as were those with round features, suggesting that minimal features that evoke “humanness” are important for establishing trust. When eyes contained red, however, faces were rated much more negatively. Color schemas appear to override anthropomorphic schemas of humanness when conflicting cues are present. Implications for the design of intelligent agents are discussed.

Anthropomorphism of Textured “Faces”

Participants rated machine “faces” which varied in terms of facial feature shape, face shape, and eight facial background textures. Ratings were made for aggression, friendliness, intelligence, trustworthiness, and degree of animation. In addition, reaction time was collected for all ratings. Rough metal and blank facial backgrounds were perceived most trustworthy. Rough metal also had the highest mean friendliness. However, across ratings, face shape and feature shape proved to have more predictive validity than did the materials making up the face. It is likely that when faced with ambiguous objects, such as the front of a novel military vehicle, people project anthropomorphic features and then make judgments accordingly.

The Effectiveness of Transferring Multi-Agent Behaviors From a Learning Environment in the Presence of Synthetic Social Structures

The diverse behavior representation schemes and learning paradigms being investigated within the robotics community share the common feature that successful deployment of agents requires that behaviors developed in a learning environment are successfully applied to a range of unfamiliar and potentially more complex operational environments. The intent of our research is to develop insight into the factors facilitating successful transfer of behaviors to the operational environments. We present experimental results investigating the effects of several factors for a simulated swarm of autonomous vehicles. Our primary focus is on the impact of Synthetic Social Structures, which are guidelines directing the interactions between agents, much like social behaviors direct interactions between group members in the human and animal world. The social structure implemented is a dominance hierarchy, which has been shown previously to facilitate negotiation between agents. The goal of this investigation is to investigate mechanisms adding robustness to agent behavior.