AIBO Research Articles

Dynamic system linear models and Bayes classifier for time series classification in promoting sustainabilitys

Research purpose: The current work introduces a novel method for time series discriminant analysis (DA). Proposing a version for the Bayes classifier employing Dynamic Linear Models, which we denote by BCDLM This article explores the application of DLMs and the Bayes Classifier in time series classification to promote application in sustainability across diverse sectors. Method: This paper presents some computer simulation studies in which we generate four different scenarios corresponding to time series observations from various Dynamic Linear Models (DLMs). In Discriminant Analysis, we investigated strategies for estimating variance in models and compared the performance of the BCDLM with other common classifiers. Such datasets are composed of real-time series (data from SONY AIBO Robot and spectrometry of coffee types) and pseudo-time series (data from Swedish leaves adapted for time series). We also point out that algorithm was used to determine training and test sets in real-world applications. Results: Considering the real-time series examined in this paper, The results obtained indicate that the parametric approach developed represents a promising alternative for this class of DA problems, with observations of time series in a situation that is quite difficult in practice when we have series with large sizes with respect to the number of observations in the classes, even though more thorough studies are required. Conclusions: It concludes that the BCDLM performed comparably to the results of the classifiers 1NN, RDA, NBND and NBK and superior to the methods LDA and QDA. This offers a powerful combination for time series classification, enabling accurate predictions and informed decision-making in areas such as energy consumption, waste management, and resource allocation.

Exploratory analysis of Sony AIBO users

It is important to understand how the cultural background, the age and the gender influence the expectations towards social robots. Although past works studied the user adaptation for some months, the users with multiple years of ownership (heavy users) were not subjects of any experiment to compare these criteria over the years. This exploratory research examines the owners of the discontinued Sony AIBO because these robots have not been abandoned by some enthusiastic users and they are still resold on the secondhand market. 78 Sony AIBO owners were recruited online and their quantitative data were analyzed by four independent variables (age, gender, culture, and length of ownership), user contribution and model preference points of view. The results revealed the motives to own these robots for years and how the heavy users perceived their social robots after a long period in the robot acceptance phase.

Improvements in the native development environment for Sony AIBO

The entertainment robotics have been on a peak with AIBO, but this robot brand has been discontinued by the Sony in 2006 to help its financial position. Among other reasons, the robot failed to enter into both the mainstream and the robotics research labs besides the RoboCup competitions, however, there were some attempts to use the robot for rehabilitation and emotional medical treatments. A native software development environment (Open-R SDK) was provided to program AIBO, nevertheless, the operating system (Aperios) induced difficulties for the students and the researchers in the software development. The author of this paper made efforts to update the Open-R and overcome the problems. More enhancements have been implemented in the core components, some software methodologies were applied to solve a number of restrictions and the achievements are summarized here.

Eliciting caregiving behavior in dyadic human-robot attachment-like interactions

We present here the design and applications of an arousal-based model controlling the behavior of a Sony AIBO robot during the exploration of a novel environment: a children's play mat. When the robot experiences too many new perceptions, the increase of arousal triggers calls for attention towards its human caregiver. The caregiver can choose to either calm the robot down by providing it with comfort, or to leave the robot coping with the situation on its own. When the arousal of the robot has decreased, the robot moves on to further explore the play mat. We gathered results from two experiments using this arousal-driven control architecture. In the first setting, we show that such a robotic architecture allows the human caregiver to influence greatly the learning outcomes of the exploration episode, with some similarities to a primary caregiver during early childhood. In a second experiment, we tested how human adults behaved in a similar setup with two different robots: one “needy”, often demanding attention, and one more independent, requesting far less care or assistance. Our results show that human adults recognise each profile of the robot for what they have been designed, and behave accordingly to what would be expected, caring more for the needy robot than for the other. Additionally, the subjects exhibited a preference and more positive affect whilst interacting and rating the robot we designed as needy. This experiment leads us to the conclusion that our architecture and setup succeeded in eliciting positive and caregiving behavior from adults of different age groups and technological background. Finally, the consistency and reactivity of the robot during this dyadic interaction appeared crucial for the enjoyment and engagement of the human partner.

Technological Nature: Adaptation and the Future of Human Life

319 Children, Youth and Environments Vol. 22 No. 1 (Spring 2012) ISSN: 1546-2250 Technological Nature: Adaptation and the Future of Human Life Kahn, Peter H. (2011). Cambridge, MA: MIT Press; 240 pages. $24.95. ISBN 0262113228. Having grown up in the heart of Manhattan, I can't say that I feel terribly expert in how people relate to nature. Indeed, during my childhood, trips to Central Park were about as far as I got from a human-engineered environment—and of course, as urban historians could attest, Central Park is itself a rather energetically "engineered" space, with carefully placed playgrounds, meadows, and ponds. For me, a typical bird is a pigeon; a typical mammal, a gray squirrel; and the insect kingdom is represented by the common cockroach. In this sense, I am a victim of "environmental generational amnesia," a condition named by Peter Kahn in his provocative bookTechnological Nature. Environmental generational amnesia, in Kahn's view, is not so much an individual as a cultural disorder, in which successive generations of human beings progressively lose more and more of their sense of the natural world; for each generation, their own experience of nature becomes the (increasingly impoverished) benchmark against which that of the still-more-diminished next generation is measured. Kahn's book is a thoughtful attempt to map out the shifting territory of how technology impacts humanity's relationship with the natural world. At the same time, it is in many important ways frustratingly incomplete. While I would strongly recommend the book to readers interested in the impact and meaning of technology—this should be a large audience—I would also stop short of blurb-level (“triumph,” “masterpiece”) language. The book raises more questions than it can possibly answer, and it left this reader, at least, with a nagging feeling that in fact not all the right questions were asked. Most of Technological Nature focuses on a series of empirical studies by Kahn and his colleagues, comparing various forms of 320 technological experience with their natural counterparts. For instance, two early chapters in the book describe experiments in which a "technological nature window" is placed in an office setting: the nature window is a large computer screen displaying, in real time, a natural (or at least pleasant outdoor) setting. In comparison to a true window looking onto the outside world, the nature window turns out to be only a pale, though not entirely valueless, substitute. While people in the presence of a real window show various physiological measures of stress relief (a finding that Kahn notes is reinforced by previous studies of windows in hospital settings), the technological nature window is no more effective at relieving stress than a blank wall. Still, for people who are ordinarily stuck in stifling indoor offices, having a nature window in their workspace for a protracted time did bring some satisfaction: at least several study participants noted that they would look at the nature window to take a mental break ("I can now find some peace in the room that wasn't there before," said one), and missed its presence after the study was over and the screen was removed. Kahn summarizes this sort of result by saying that "technological nature is better than no nature but not as good as actual nature" (137). In a similar vein, he recounts several studies in which children interact with the Sony AIBO robotic dog; in one study, preschool children's responses to AIBO are compared to their responses to a stuffed dog; in another, the responses of somewhat older children are compared to interactions with a real (biological) dog. There are many fascinating observations in the book regarding these experiments, but the bottom line is that the preschool children clearly treat AIBO with more interest and "lifelike" care than a stuffed dog, while the older children are much more affectionate and tactile with the biological dog than with AIBO. Again, the robot dog is in a space between inanimate and animate— in Kahn's phrasing, better than no nature but not as good as actual nature. Interestingly, in yet another chapter, Kahn describes encouraging results in which autistic children have far more social responses...

Learning from Demonstration in Spatial Exploration

We present the initial stage of our research on Learning from Demonstration algorithms. We have implemented an algorithm based on Confident Execution, one of the components of the Confidence-Based Autonomy algorithm developed by Chernova and Veloso. Our preliminary experiments were conducted first in simulation and then using a Sony AIBO ERS-7 robot. So far, our robot has been able to learn crude navigation strategies, despite limited trials. We are currently working on improving our implementation by including additional features that describe more broadly the state of the agent. Our long term goal is to incorporate Learning from Demonstration techniques in our HRTeam (human/multi-robot) framework.

206102 四脚移動ロボットの視覚情報を利用した認識性能の向上(一般02 機械力学・制御工学4)

This paper presents how to improve the visual recognition ability of SONY AIBO robot. AIBO is a general purpose quadruped robot and most users enjoy the robot through interactions with AIBO by means of a variety of sensors mounted thereon. The sensors include CMOS camera, infrared ranging sensor, microphone, accelerometer, many touch sensors and so on. This study focuses on use of the CMOS camera. Recognition task of the camera is sharply affected by ambient lighting conditions. In our approach, a shutter speed is adjusted so as to improve color recognition ability under different lighting conditions. The approach is verified through experiments under a variety of conditions.

Experience in Developing a Robot Control Software

This article presents an experience report in using the UML-driven development process to develop an object-oriented control software for a Sony AIBO robot. The entire project was a great learning experience for all the team members and we found the methodologies we used very effective in helping us design and develop a high-quality control system. We also feel that other software developers and practitioners, especially those developing robotic and embedded control software, would benefit from the experience and observations reported in this article.

Confidence-Based Multi-Robot Learning from Demonstration

Learning from demonstration algorithms enable a robot to learn a new policy based on demonstrations provided by a teacher. In this article, we explore a novel research direction, multi-robot learning from demonstration, which extends demonstration based learning methods to collaborative multi-robot domains. Specifically, we study the problem of enabling a single person to teach individual policies to multiple robots at the same time. We present flexMLfD, a task and platform independent multi-robot demonstration learning framework that supports both independent and collaborative multi-robot behaviors. Building upon this framework, we contribute three approaches to teaching collaborative multi-robot behaviors based on different information sharing strategies, and evaluate these approaches by teaching two Sony QRIO humanoid robots to perform three collaborative ball sorting tasks. We then present scalability analysis of flexMLfD using up to seven Sony AIBO robots. We conclude the article by proposing a formalization for a broader multi-robot learning from demonstration research area.

Genetic Programming-Based Automatic Gait Generation in Joint Space for a Quadruped Robot

This paper introduces a new approach to developing a fast gait for a quadruped robot using genetic programming (GP). Planning gaits for legged robots is a challenging task that requires optimizing parameters in a highly irregular and multi-dimensional space. Several recent approaches have focused on using genetic algorithms (GAs) to generate gaits automatically and have shown significant improvement over previous gait optimization results. Most current GA-based approaches optimize only a small, pre-selected set of parameters, but it is difficult to decide which parameters should be included in the optimization to get the best results. Moreover, the number of pre-selected parameters is at least 10, so it can be relatively difficult to optimize them, given their high degree of interdependence. To overcome these problems of the typical GA-based approach, we have proposed a seemingly more efficient approach that optimizes joint trajectories instead of locus-related parameters in Cartesian space, using GP. Our GP-based method has obtained much-improved results over the GA-based approaches tested in experiments on the Sony AIBO ERS-7 in the Webots environment. The elite archive mechanism is introduced to combat the premature convergence problems in GP and has shown better results than a traditional multi-population approach.

Actor-Critic Learning for Platform-Independent Robot Navigation

This article describes an approach in the field of reinforcement learning for robot control and a new Modular Actor-Critic architecture which supports platform-independent robot control. The architecture is tested on a landmark approaching task using movable pan/tilt cameras which successfully control both a large PeopleBot and a small Sony Aibo robot to perform the navigation task, with no retraining required. The architecture provides insight into the skills transfer between different robotic platforms and the modularisation of the architecture derived from splitting the control tasks into their component parts. The architecture and underlying principles could be used in rapid prototyping of new robotic platforms, where an already functioning control system can be used to allow more sophisticated navigation.

Children's behavior toward and understanding of robotic and living dogs

This study investigated children's reasoning about and behavioral interactions with a computationally sophisticated robotic dog (Sony's AIBO) compared to a live dog (an Australian Shepherd). Seventy-two children from three age groups (7–9 years, 10–12 years, and 13–15 years) participated in this study. Results showed that more children conceptualized the live dog, as compared to AIBO, as having physical essences, mental states, sociality, and moral standing. Children also spent more time touching and within arms distance of the live dog, as compared to AIBO. However, a surprising majority of children conceptualized and interacted with AIBO in ways that were like a live dog. For example, over 60% of the children affirmed that AIBO had mental states, sociality, and moral standing; and children were as likely to give AIBO commands as a living dog. Discussion broaches whether it is possible that a new technological genre is emerging that challenges traditional ontological categories.

A hybrid approach to fast and accurate localization for legged robots

SUMMARYThis paper describes a hybrid approach to a fast and accurate localization method for legged robots based on Fuzzy-Markov (FM) and Extended Kalman Filters (EKF). Both FM and EKF techniques have been used in robot localization and exhibit different characteristics in terms of processing time, convergence, and accuracy. We propose a Fuzzy-Markov–Kalman (FM–EKF) localization method as a combined solution for a poor predictable platform such as Sony Aibo walking robots. The experimental results show the performance of EKF, FM, and FM-EKF in a localization task with simple movements, combined behaviors, and kidnapped situations. An overhead tracking system was adopted to provide a ground truth to verify the performance of the proposed method.

An appearance-based visual compass for mobile robots

Localization is one of the most important basic skills of a mobile robot. Most approaches, however, still rely either on special sensors or require artificial environments. In this article, a novel approach is presented that can provide compass information for localization, purely based on the visual appearance of a room. A robot using such a visual compass can quickly learn a cylindrical map of the environment, consisting of simple statistical features that can be computed very quickly. The visual compass algorithm is efficient, scalable and can therefore be used in real-time on almost any contemporary robotic platform. Extensive experiments on a Sony Aibo robot have validated that the approach works in a vast variety of environments.

GA를 이용한 4족 보행로봇의 걸음새 자동 생성 및 성능향상

GA 기반의 4족 보행로봇의 걸음새의 속도와 안정성을 최적화하는 걸음새의 자동 생성 방법을 구현한다. 기존에 사용된 걸음새 파라미터 집합에서 중요 파라미터의 영향을 분석하였고, 이를 통해 효율적인 탐색 방향을 설정하였다. 또한 속도 위주의 기존 연구와는 달리 반 정확도가 수반된 결과를 얻도록 하였다. 제안된 기법의 검증을 위하여 SONY Aibo 4족 보행 로봇에 대해서 ODE 기반의 물리적 특성을 포함한 정교한 시뮬레이션이 가능한 Webots 을 이용하여 실험을 수행하였고, 속도와 안정성 면에서 향상된 결과를 얻었다. This Paper introduces new approach to develop fast and reliable gaits for quadruped robot using GA(genetic algorithm). Planning gaits for legged robots is a challenging task that requires optimizing parameters in a highly irregular and multidimensional space. Recent approaches have problems to select proper parameters which are not known in advance and optimize more than ten to twenty parameters simultaneously. In our approach, the effects of major gait parameters are analysed and used to guide the search more efficiently. The experiments of Sony AIBO ERS-7 in Webots environment indicate that our approach is able to produce much improved results in fast velocity and reliability.

관절 공간에서의 GP 기반 진화기법을 이용한 4족 보행로봇의 걸음새 자동생성

This paper introduces a new approach to develop a fast gait for quadruped robot using GP(genetic programming). Planning gaits for legged robots is a challenging task that requires optimizing parameters in a highly irregular and multidimensional space. Several recent approaches have focused on using GA(genetic algorithm) to generate gait automatically and shown significant improvement over previous results. Most of current GA based approaches used pre-selected parameters, but it is difficult to select the appropriate parameters for the optimization of gait. To overcome these problems, we proposed an efficient approach which optimizes joint angle trajectories using genetic programming. Our GP based method has obtained much better results than GA based approaches for experiments of Sony AIBO ERS-7 in Webots environment.

A Prey Catching and Predator Avoidance Neural-Schema Architecture for Single and Multiple Robots

The paper presents a biologically inspired multi-level neural-schema architecture for prey catching and predator avoidance in single and multiple autonomous robotic systems. The architecture is inspired on anuran (frogs and toads) neuroethological studies and wolf pack group behaviors. The single robot architecture exploits visuomotor coordination models developed to explain anuran behavior in the presence of preys and predators. The multiple robot architecture extends the individual prey catching and predator avoidance model to experiment with group behavior. The robotic modeling architecture distinguishes between higher-level schemas representing behavior and lower-level neural structures representing brain regions. We present results from single and multiple robot experiments developed using the NSL/ASL/MIRO system and Sony AIBO ERS-210 robots.

Applying the Appraisal Theory of Emotion to Human-Agent Interaction

When people interact with one another, there is a series of conscious and unconscious evaluations used to judge the situation in order to determine an emotional response. This research examines whether the emotional appraisals that individuals use when interacting with other humans, can be applied to human-agent interactions, and whether the attributes of the non-human agent affect the nature of these appraisals. Participants work with one of three non-human teammates to accomplish a series of tasks. These agents are a real dog, a robotic dog (Sony AIBO), and a nondescript robot (Lego NXT). Participants' emotional reaction is measured through subjective questionnaires, physiological data (EKG & galvanic skin response), and vocal analysis. Taken together this set of measurements forms a detailed picture of how humans react emotionally to agents during their task interaction. It is predicted that agent form will influence participants' appraisals and emotional reactions.

Caveat Venditor: Technologically Protected Subsidized Goods and the Customers Who Hack Them

This paper explores the issues surrounding the subsidization of a technology based durable good and on the delicate dance between the producer trying to protect their profit, competitors trying to to create and sell aftermarket goods, and innovative customers who attempt to use the goods in completely unplanned, and unprofitable ways. A number of case studies are presented that highlighted the ease with which customers can tinker with subsidized products. These include Microsoft's Xbox, Netpliances' i-Opener, the Sony AIBO, and the general problems of the prepaid phone industry. A number of legal cases are also presented, although these focus more on the issue of other firms attempting to make competing aftermarket products for subsidized primary goods. The difficult question of What can a company do is explored. The question of who the company can go after for infringement, be it the reverse engineering programmer, or the college student sharing such information on her homepage, is also examined. Finally, this paper examines the moral issues involved when consumers hack these items.

Machine Learning With AIBO Robots in the Four-Legged League of RoboCup

Robot learning is a growing area of research at the intersection of robotics and machine learning. The main contributions of this paper include a review of how machine learning has been used on Sony AIBO robots and at RoboCup, with a focus on the four-legged league during the years 1998-2004. The review shows that the application-oriented use of machine learning in the four-legged league was still conservative and restricted to a few well-known and easy-to-use methods such as standard decision trees, evolutionary hill climbing, and support vector machines. Method-oriented spin-off studies emerged more frequently and increasingly addressed new and advanced machine learning techniques. Further, the paper presents some details about the growing impact of machine learning in the software system developed by the authors' robot soccer team-the NUbots