Adaptive behaviors of reactive mobile robot with Bayesian inference in nonstationary environments

This paper presents a technique for a reactive mobile robot to adaptively behave in unforeseen and dynamic circumstances. A robot in nonstationary environments needs to infer how to adaptively behave to the changing environment. Behavior-based approach manages the interactions between the robot and its environment for generating behaviors, but in spite of its strengths of fast response, it has not been applied much to more complex problems for high-level behaviors. For that reason many researchers employ a behavior-based deliberative architecture. This paper proposes a 2-layer control architecture for generating adaptive behaviors to perceive and avoid moving obstacles as well as stationary obstacles. The first layer is to generate reflexive and autonomous behaviors with behavior network, and the second layer is to infer dynamic situations of the mobile robot with Bayesian network. These two levels facilitate a tight integration between high-level inference and low-level behaviors. Experimental results with various simulations and a real robot have shown that the robot reaches the goal points while avoiding stationary or moving obstacles with the proposed architecture.     

The development of an autonomous service robot. Implementation: “Lucas”—The library assistant robot

This paper describes an autonomous mobile device that was designed, developed and implemented as a library assistant robot. A complete autonomous system incorporating human–robot interaction has been developed and implemented within a real world environment. The robotic development is comprehensively described in terms of its localization systems, which incorporates simple image processing techniques fused with odometry and sonar data, which is validated through the use of an extended Kalman filter (EKF). The essential principles required for the development of a successful assistive robot are described and put into demonstration through a human–robot interaction application applied to the library assistant robot.     

Towards an intelligent system for generating an adapted verbal and nonverbal combined behavior in human–robot interaction

In human–robot interaction scenarios, an intelligent robot should be able to synthesize an appropriate behavior adapted to human profile (i.e., personality). Recent research studies discussed the effect of personality traits on human verbal and nonverbal behaviors. The dynamic characteristics of the generated gestures and postures during the nonverbal communication can differ according to personality traits, which similarly can influence the verbal content of human speech. This research tries to map human verbal behavior to a corresponding verbal and nonverbal combined robot behavior based on the extraversion–introversion personality dimension. We explore the human–robot personality matching aspect and the similarity attraction principle, in addition to the different effects of the adapted combined robot behavior expressed through speech and gestures, and the adapted speech-only robot behavior, on interaction. Experiments with the humanoid NAO robot are reported.     

Development of “STORK”, a watermelon-harvesting robot

Recently, the production of heavy fruit and vegetables has been decreasing in Japan because strenuous labor is require to harvest them. A robot would allow them to be harvested more easily. We have developed the robot “STORK” to harvest watermelons. STORK has a low mass and a long working range. The position accuracy and repeatability of the manipulator, the required vacuum, and the allowance for position error of the vacuum pad were tested.     

Effect of robot’s gaze behaviors for attracting and controlling human attention

Controlling someone’s attention can be defined as shifting his/her attention from the existing direction to another. To shift someone’s attention, gaining attention and meeting gaze are two most important prerequisites. If a robot would like to communicate a particular person, it should turn its gaze to him/her for eye contact. However, it is not an easy task for the robot to make eye contact because such a turning action alone may not be effective in all situations, especially when the robot and the human are not facing each other or the human is intensely attending to his/her task. Therefore, the robot should perform some actions so that it can attract the target person and make him/her respond to the robot to meet gaze. In this paper, we present a robot that can attract a target person’s attention by moving its head, make eye contact through showing gaze awareness by blinking its eyes, and directs his/her attention by repeating its eyes and head turns from the person to the target object. Experiments using 20 human participants confirm the effectiveness of the robot actions to control human attention.     

Systems, control models, and codec for collaborative observation of remote environments with an autonomous networked robotic camera

Networked robotic cameras are becoming popular in remote observation applications such as natural observation, surveillance, and distance learning. Equipped with a high optical zoom lens and agile pan-tilt mechanisms, a networked robotic camera can cover a large region with various resolutions. The optimal selection of camera control parameters for competing observation requests and the on-demand delivery of video content for various spatiotemporal queries are two challenges in the design of such autonomous systems. For camera control, we introduce memoryless and temporal frame selection models that effectively enable collaborative control of the camera based on the competing inputs from in-situ sensors and users. For content delivery, we design a patch-based motion panorama representation and coding/decoding algorithms (codec) to allow efficient storage and computation. We present system architecture, frame selection models, user interface, and codec algorithms. We have implemented the system and extensively tested our design in real world applications including natural observation, public surveillance, distance learning, and building construction monitoring. Experiment results show that our frame selection models are robust and effective and our on-demand content delivery codec can satisfy a variety of spatiotemporal queries efficiently in terms of computation time communications bandwidth.

SEDRET—an intelligent system for the diagnosis and prediction of events in power plants

Artificial Intelligence applications in large-scale industry, such as fossil power plants, require the ability to manage uncertainty and time. In this paper, we present an intelligent system to assist an operator of a power plant. This system, called SEDRET, is based on a novel knowledge representation of uncertainty and time, called Temporal Nodes Bayesian Networks (TNBN), a type of Probabilistic Temporal Network. A set of temporal nodes and a set of edge define a TNBN, each temporal node is defined by a value of a variable and a time interval associate to the change of variable value. A TNBN generates a formal and systematic structure for modeling the temporal evolution of a process under uncertainty. The inference mechanism is based on probabilistic reasoning. A TNBN can be used to recognize events and state variables with respect to current plant conditions and predict the future propagation of disturbances. SEDRET was validated with the diagnosis and prediction of events in a steam generator with a power plant training simulator. The results performed in this work indicate that SEDRET can potentially improve plant availability through early diagnosis and prediction of disturbances that could lead to plant shutdown.     

Dynamic testing: Can a robot as tutor be of help in assessing children's potential for learning?

This study examined whether computerized dynamic testing by utilizing a robot would lead to different patterns in children's (aged 6–9 years) potential for learning and strategy use when solving series‐completion tasks. The robot, in a “Wizard of Oz” setting, provided instructions and prompts during dynamic testing. It was found that a dynamic training resulted in greater accuracy and more correctly placed pieces at the post‐test than repeated testing only. Moreover, children who were dynamically trained appeared to use more heuristic strategies at the post‐test than their peers who were not trained. In general, observations showed that children were excited to work with the robot. All in all, the study revealed that computerized dynamic testing by means of a robot has much potential in tapping into children's potential for learning and strategy use. The implications of using a robot in educational assessment were stressed further in the discussion.     

Development of the quadruped walking robot,TITAN-IX — mechanical design concept and application for the humanitarian de-mining robot

This paper proposes a quadruped walking robot that has high performance as a working machine. This robot is needed for various tasks controlled by tele-operation, especially for humanitarian mine detection and removal. Since there are numerous personnel landmines that are still in place from many wars, it is desirable to provide a safe and inexpensive tool that civilians can use to remove those mines. The authors have been working on the concept of the humanitarian demining robot systems for 4 years and have performed basic experiments with the first prototype VK-I using the modified quadruped walking robot, TITAN-VIII. After those experiments, it was possible to refine some concepts and now the new robot has a tool (end-effector) changing system on its back, so that by utilizing the legs as manipulation arms and connecting various tools to the foot, it can perform mine detection and removal tasks. Toaccomplish these tasks, we developed various end-effectors that can be attached to the working leg. In this paper we will discuss the mechanical design of the new walking robot called TITAN-IX to be applied to the new system VK-II.     

Development of an intelligent model to estimate the height of caving–fracturing zone over the longwall gobs

Neural Computing and Applications - After the ore (seam) extraction in longwall mining, the immediate roof layers over the extracted panel are strained and suspended downward. This process expands...     

Design of an H∞ controller for the Delta robot: experimental results

This paper deals with an efficient implementation of an H∞ multi-variable controller on the three degrees of freedom (DOF) parallel robot namely the ‘Delta robot’. The H∞ controller is designed by the mixed sensitivity approach in which the sensitivity function matrix S and the complementary sensitivity function matrix T are taken into account. For this purpose, a nonlinear analytical dynamic state model is developed and a tangent linearization procedure is used to obtain a multi-variable linear model around a functional point. Real-time experiments were performed to compare the centralized H∞ controller with a classical decentralized Proportional Integral Derivative (PID) controller. Experimental tracking results show that the performances of the PID compared to those of the H∞ decrease when the movement dynamic is increased. At high dynamic (12 Ge), it is shown that the maximum tracking error and the error around the stop positions of the H∞ are, respectively, 80 and 60% of the PID. The experiments of the load variation have proven that the H∞ is more robust than the PID. The steady-state root mean square error of the H∞ is less than 60% of the one obtained using the PID controller.     

Principles of transparency for autonomous vehicles: first results of an experiment with an augmented reality human–machine interface

Cognition, Technology & Work - Highly automated driving allows the driver to temporarily delegate the driving task to the autonomous vehicle. The challenge is to define the information that...     

Knowledge specification and representation for an “intelligent” interface devoted to process monitoring and supervision

This paper presents our approach of cooperation between a human operator and a decision aid tool by means of an “intelligent” interface manager: the Decisional Module of Imagery (D.M.I.). The “heart” of the D.M.I. is an expert system which manipulates three main objects (the WHAT, WHEN and HOW objects) described next. Knowledge specification and representation for the expert system, and the way we chose to implement it, are then explained.     

Robots in the classroom: Differences in students’ perceptions of credibility and learning between “teacher as robot” and “robot as teacher”

Advancements in technology are bringing robotics into interpersonal communication contexts, including the college classroom. This study was one of the first to examine college students’ communication-related perceptions of robots being used in an instructional capacity. Student participants rated both a human instructor using a telepresence robot and an autonomous social robot delivering the same lesson as credible. However, students gave higher credibility ratings to the teacher as robot, which led to differences between the two instructional agents in their learning outcomes. Students reported more affective learning from the teacher as robot than the robot as teacher, despite controlled instructional performances. Instructional agent type had both direct and indirect effects on behavioral learning. The direct effect suggests a potential machine heuristic in which students are more likely to follow behavioral suggestions offered by an autonomous social robot. The findings generally support the MAIN model and the Computers are Social Actors paradigm, but suggest that future work needs to be done in this area.     

A hybrid cognitive/reactive intelligent agent autonomous path planning technique in a networked-distributed unstructured environment for reinforcement learning

This paper proposes a path planning technique for autonomous agent(s) located in an unstructured networked distributed environment, where each agent has limited and not complete knowledge of the environment. Each agent has only the knowledge available in the distributed memory of the computing node the agent is running on and the agents share some information learned over a distributed network. In particular, the environment is divided into several sectors with each sector located on a single separate distributed computing node. We consider hybrid reactive-cognitive agent(s) where we use autonomous agent motion planning that is based on the use of a potential field model accompanied by a reinforcement learning as well as boundary detection algorithms. Potential fields are used for fast convergence toward a path in a distributed environment while reenforcement learning is used to guarantee a variety of behavior and consistent convergence in a distributed environment. We show how the agent decision making process is enhanced by the combination of the two techniques in a distributed environment. Furthermore, path retracing is a challenging problem in a distributed environment, since the agent does not have complete knowledge of the environment. We propose a backtracking technique to keep the distributed agent informed all the time of its path information and step count including when migrating from one node to another. Note that no node has knowledge of the entire global path from a source to a goal when such a goal resides on a separate node. Each agent has only knowledge of a partial path (internal to a node) and related number of steps corresponding to the portion of the path that agent traversed when running on the node. In particular, we show how each of the agents(s), starting in one of the many sectors with no initial knowledge of the environment, using the proposed distributed technique, develops its intelligence based on its experience and seamlessly discovers the shortest global path to the target, which is located in a different node, while avoiding any obstacle(s) it encounters in its way, including when transitioning and migrating from one distributed computing node to another. The agent(s) use (s) multiple-token-ring message passing interface (MPI) to perform internode communication. Finally, the experimental results of the proposed method show that single and multiagents sharing the same goal and running on the same or different nodes successfully coordinate the sharing of their respective environment states/information to collaboratively perform their respective tasks. The results also show that distributed multiagent sharing information increases by an order of magnitude the speed of convergence to the optimal shortest path to the goal in comparison with the single-agent case or noninformation sharing multiagent case.     

Developing a taxonomy of faculty participation in asynchronous learning environments—an exploratory investigation

Anecdotal evidence suggests that instructor performance in the online discussion portion of online courses has a major impact on learning and learner satisfaction. If this is the case, faculty development is an important component of success. This paper describes the development of a taxonomy of six categories to describe instructor postings to asynchronous discussions in online courses. This taxonomy was used during the inspection of 18 post-graduate level course sections at a Midwest university. The resulting data show that there is considerable variation in faculty teaching styles, interaction, and the amount of content-related feedback. The process also allowed issues such timeliness of feedback and responsiveness of faculty to be quantified. In-depth interviews with instructors and learners with experience of online courses were conducted to validate the qualitative data.