Z. Zenn Bien

Human-machine interface for wheelchair control with EMG and its evaluation

The objective of this paper is to develop a powered wheelchair controller based on EMG for users with high-level spinal cord injury. EMG is very naturally measured when the user indicating a certain direction and the force information which will be used for the speed of wheelchair is easily extracted from EMG. Furthermore, the emergency situation based on EMG will be checked relatively ease. We classified the pre-defined motions such as rest case, forward movement, left movement, and right movement by fuzzy min-max neural networks (FMMNN). This classification results and evaluation results with real users shows the feasibility of EMG as an input interface for powered wheelchair.

Intention reading is essential in human-friendly interfaces for the elderly and the handicapped

Some major issues of human-friendly man-machine interaction/interfaces in an intelligent residential system are dealt with to cope with increasing needs of the elderly and the handicapped for a higher level of living conditions. Specifically, the existing systems in this area are first examined and the inadequacy of some current remote-control interfaces for the elderly/handicapped is pointed out. Then, some intention-reading techniques are introduced for the elderly and/or the physically handicapped, for easier and more convenient use of machines and engineering systems.

Iterative learning controllers for discrete-time large-scale systems to track trajectories with distinct magnitudes

In the procedure of steady-state hierarchical optimization for large-scale industrial processes, it is often necessary that the control system responds to a sequence of step function-type control decisions with distinct magnitudes. In this paper a set of iterative learning controllers are de-centrally embedded into the procedure of the steady-state optimization. This generates upgraded sequential control signals and thus improves the transient performance of the discrete-time large-scale systems. The convergence of the updating law is derived while the intervention from the distinction of the scales is analysed. Further, an optimal iterative learning control scheme is also deduced by means of a functional derivation. The effectiveness of the proposed scheme and the optimal rule is verified by simulation.

Soft Computing Based Emotion/Intention Reading for Service Robot

Due to its tolerance to imprecision, uncertainty and partial truth, the soft computing technique deals well with human related signals such as voice, gesture, facial expression, bio-signal, etc. In this paper, we propose architecture of soft computing based recognition for a class of biosign. Especially, the problem of inferring emotion and intention reading from such recognition is considered with applications for service robots that interact with human. It is shown that proposed architecture renders good performance in a few experimental systems for rehabilitation.

Development of a steward robot for human-friendly interaction

Independence of people in need in their living environment becomes an imperative notion for any society in the years to come. As an approach to achieve independence, this paper addresses the problem of controlling an assistive home environment emphasizing a human-friendly man-machine interaction. We introduce a new service robot, named Joy, which can be categorized as a steward robot. As an intermediate agent between inhabitants and a complex smart house environment, this robot can provide the users with easily accessible, convenient, and cost effective environment for independent living. The learning capability and emotional interaction of the robot can enhance human-friendliness in various tasks. A learning system can provide personalized services by accumulating knowledge of the behavior patterns of the user in daily activities. An emotional interaction system generates facial expressions to be more human-friendly. We have developed two types of steward robots: a software type and hardware type. The hardware-type robot has a mobile platform and two robotic arms to perform various tasks with mobility.

LEARNING TECHNIQUES IN SERVICE ROBOTIC ENVIRONMENT

This presentation addresses the problems of realizing human-friendly man-machine interaction in service robotic environment with emphasis on learning capability. After briefly reviewing the issues of human-robot interaction and various learning techniques from engineering point of view, we report our experiences in case studies where some learning techniques are successfully implemented in service robotic environment and discuss open issues of learning systems such as adaptivity and life-long learning capability.

Personalized Media Control Method using Probabilistic Fuzzy Rule-based Learning

Intention reading technique is essential to provide personalized services toward more convenient and human-friendly services in complex ubiquitous environment such as a smart home. If a system has knowledge about an user`s intention of his/her behavioral pattern, the system can provide mote qualified and satisfactory services automatically in advance to the user`s explicit command. In this sense, learning capability is considered as a key function for the intention reading technique in view of knowledge discovery. In this paper, ore introduce a personalized media control method for a possible application iii a smart home. Note that data pattern such as human behavior contains lots of inconsistent data due to limitation of feature extraction and insufficiently available features, where separable data groups are intermingled with inseparable data groups. To deal with such a data pattern, we introduce an effective engineering approach with the combination of fuzzy logic and probabilistic reasoning. The proposed learning system, which is based on IFCS (Iterative Fuzzy Clustering with Supervision) algorithm, extract probabilistic fuzzy rules effectively from the given numerical training data pattern. Furthermore, an extended architectural design methodology of the learning system incorporating with the IFCS algorithm are introduced. Finally, experimental results of the media contents recommendation system are given to show the effectiveness of the proposed system.

Similarity between Fuzzy Multi-objective Control and Eligibility

A fuzzy multi-objective control problem has been handled in many different ways such as neural network and reinforcement learning etc. Among them, reinforcement learning solves a fuzzy multi-objective control problem without any priori knowledge about an environment. In this paper, a new method of reinforcement learning for a fuzzy multi-objective control problem is proposed in consideration of newly defined objective TD(?), where TD stands for a temporal difference. The proposed method reformulates a fuzzy multi-objective control problem into a problem similar to a reinforcement learning problem under non-Markov environment, where objective eligibility is considered for handling multi-rewards, similarly as TD(?) is applied to a reinforcement learning problem under a non-Markov environment.