In this study, the authors propose a novel three‐dimensional (3D) convolutional neural network for shape reconstruction via a trilateral convolutional neural network (Tri‐CNN) from a single depth view. The proposed approach produces a 3D voxel representation of an object, derived from a partial object surface in a single depth image. The proposed Tri‐CNN combines three dilated convolutions in 3D to expand the convolutional receptive field more efficiently to learn shape reconstructions. To evaluate the proposed Tri‐CNN in terms of reconstruction performance, the publicly available ShapeNet and Big Data for Grasp Planning data sets are utilised. The reconstruction performance was evaluated against four conventional deep learning approaches: namely, fully connected convolutional neural network, baseline CNN, autoencoder CNN, and a generative adversarial reconstruction network. The proposed experimental results show that Tri‐CNN produces superior reconstruction results in terms of intersection over union values and Brier scores with significantly less number of model parameters and memory.

This paper introduces a humanoid robot system for evaluating human-robot interaction. A humanoid robot. Kibo. is designed to walk and dance with 22 joints and specially has a robot face to express its emotion by moving lips and eyebrows, etc. To evaluate emotional states of a human operator, face and gesture expressions are recognized by external stereo vision system and speech commands are recognized with a wireless microphone. On the view point of software development. Intelligent Robot Software Architecture developed by Center for Intelligent Robotics is adopted. Since the software architecture is designed for reusability and extensibility, rapid S/W development was achieved. As primary experiments. Kibo imitates human operator's facial and gesture expression.

Home service robots increasingly need to provide diverse and complex services such as cooking, sweeping and dishwashing. These services inevitably require a number of software functions simultaneously. For example, the cooking service requires an arm manipulation function to grasp dishes, an navigation function to move around, an object recognition function to find foods, an speech recognition function to understand user requirements, and etc. However, when the services and software functions are executed simultaneously in a robot without run-time software management, those may cause malfunction due to resource contention. In this paper, we describe the situation that causes resource contention and formulate architecture-based adaptation in robot software systems. Based on the formulation we proposed an approach to dynamic robot software management that effectively uses robot computing resources.

This paper introduces the system development of the ubiquitous computing environment, called easy interaction room (EIR), in which humans can get intelligent robotic services with easy interaction. The room has been built to research the roles between the ubicomp environment and service robot in it. Also EIR system collaborates on a special home service with the monolithic robot platform. In this development, cutting-edge technologies are employed on the design and implementation of hardware infrastructure for the automated environment. Intelligent Robot Software Architecture, developed by CIR for reusable and extensible architecture, is used to build EIR's software architecture for rapid development. Preliminary correlations between EIR and a service robot have been designed and implemented on the architectures.

The research field of Intelligent Service Robots, which has become more and more popular over the last years, covers a wide range of applications from climbing machines for cleaning large storefronts to robotic assistance for disabled or elderly people. When developing service robot software, it is a challenging problem to design the robot architecture by carefully considering user needs and requirements, implement robot application components based on the architecture, and integrate these components in a systematic and comprehensive way for maintainability and reusability. Furthermore, it becomes more difficult to communicate among development teams and with others when many engineers from different teams participate in developing the service robot. To solve these problems, we applied the COMET design method, which uses the industry-standard UML notation, to developing the software of an intelligent service robot for the elderly, called T-Rot, under development at Center for Intelligent Robotics (CIR). In this paper, we discuss our experiences with the project in which we successfully addressed these problems and developed the autonomous navigation system of the robot with the COMET/UML method.