As the development of autonomous vehicles becomes realistic, many automobile manufacturers and components producers aim to develop ‘completely autonomous driving’. ADAS (Advanced Driver Assistance Systems) which has been applied in automobile recently, supports the driver in controlling lane maintenance, speed and direction in a single lane based on limited road environment. Although technologies of obstacles avoidance on the obstacle environment have been developed, they concentrates on simple obstacle avoidances, not considering the control of the actual vehicle in the real situation which makes drivers feel unsafe from the sudden change of the wheel and the speed of the vehicle. In order to develop the ‘completely autonomous driving’ automobile which perceives the surrounding environment by itself and operates, ability of the vehicle should be enhanced in a way human driver does. In this sense, this paper intends to establish a strategy with which autonomous vehicles behave human-friendly based on vehicle dynamics through the reinforcement learning that is based on Q-learning, a type of machine learning. The obstacle avoidance reinforcement learning proceeded in 5 simulations. The reward rule has been set in the experiment so that the car can learn by itself with recurring events, allowing the experiment to have the similar environment to the one when humans drive. Driving Simulator has been used to verify results of the reinforcement learning. The ultimate goal of this study is to enable autonomous vehicles avoid obstacles in a human-friendly way when obstacles appear in their sight, using controlling methods that have previously been learned in various conditions through the reinforcement learning.

The development of a face robot basically targets very natural human-robot interaction (HRI), especially emotional interaction. So does a face robot introduced in this paper, named Buddy. Since Buddy was developed for a mobile service robot, it doesn’t have a living-being like face such as human’s or animal’s, but a typically robot-like face with hard skin, which maybe suitable for mass production. Besides, its structure and mechanism should be simple and its production cost low. This paper introduces the mechanisms and functions of mobile face robot named Buddy which can take on natural and precise facial expressions and make dynamic gestures driven by one laptop PC. Buddy also can perform lip-sync, eye-contact, face-tracking for lifelike interaction. In addition, by adopting a customized emotional reaction decision model, Buddy can create own personality, emotion and motive using various sensor data input. Based on this model, Buddy can interact probably with users and perform real-time learning using personality factors. The interaction performance of Buddy is successfully demonstrated by experiments and simulations.

본 논문에서는 인간 운동 제어 이론과 기계학습을 기반으로 하여 로봇의 접촉 작업 수행을 위한 새로운 운동 학습 전략을 제시하였다. 성공적인 접촉 작업 수행을 위한 본 연구의 전략은 강화학습 기법을 통하여 최적의 작업 수행을 위한 임피던스 매개 변수를 찾는 것이다. 본 연구에서는 최적의 임피던스 매개 변수를 결정하기 위하여 Recursive Least-Square (RLS) 필터 기반 episodic Natural Actor-Critic 알고리즘이 적용되

This study developed a novel augmented reality interface for minimally invasive surgery. The augmented reality technique can alleviate the sensory feedback problem inherent to laparoscopic surgery. An augmented reality system merges real laparoscope image and reconstructed 3D patient model based on diagnostic medical image such as CT, MRI data. By using reconstructed 3D patient model, AR interface could express structure of patient body that is invisible outside visual field of laparoscope. Therefore, an augmented reality system improved sight information of limited laparoscope. In our augmented reality system, the laparoscopic view is located at the center of a wide-angle concave screen and reconstructed 3D patient model is displayed outside the laparoscope. By using a joystick, the laparoscopic view and the reconstructed 3D patient model view are changed concurrently. With our augmented reality system, the surgeon can see the peritoneal cavity from a wide angle of view, without having to move the laparoscope. Since the concave screen serves immersive environments, the surgeon can feel as if she is in the patient body. For these reasons, a surgeon can recognize easily depth information about inner parts of patient and position information of surgical instruments without laparoscope motion. It is possible for surgeon to manipulate surgical instruments more exact and fast. Therefore immersive augmented reality interface for minimally invasive surgery will reduce bodily, environmental load of a surgeon and increase efficiency of MIS.

Unlike robotic systems, humans excel at a variety of tasks by utilizing their intrinsic impedance, force sensation, and tactile contact clues. By examining human strategy in arm impedance control, we may be able to teach robotic manipulator's human's superior motor skills in contact tacks.This paper develops a novel method for estimating and predicting the human joint impedance using the electromyogram(EMG)signals and limb position measurements. The EMG signal is the summation of MUAPs(motor unit action potentials). Determination of the relationship between the EMG signals and joint stiffness is difficult, due to irregularities and uncertainties of the EMG signals. In this research, an artificial neural network(ANN)model was developed to model the relation between the EMG and joint stiffness. The proposed method estimates and predicts the multi joint stiffness without complex calculation and specialized apparatus. The feasibility of the developed model was confirmed by experiments and simulations.

In a number of fields, robots are being used for two purposes: efficiency and safety. Most robots, however, have single-actuator mechanism for each joint, where the tasks are performed with high stiffness. High stiffness causes undesired problems to the environment and robots. This study proposes redundant actuator mechanism as an alternative idea to cope with these problems. In this paper, Double-Actuator Unit (DAU) is implemented at each joint for applications of multi-link manipulators. The DAU is composed of two motors: the positioning actuator and the stiffness modulator, which enables independent control of positioning and compliance. A three-link manipulator with DAUs enables adaptive control of RCC. By modulating the joint stiffness of the manipulator and controlling the position of RCC, we can significantly reduce contact force during assembly tasks and surgical procedures.