Thavida Maneewarn

A modified approach to fuzzy Q learning for mobile robots

A modified approach to fuzzy Q-learning is presented in this paper. A reward sharing mechanism is added to increase the learning speed and to allow treatment of each fuzzy rule as a separate learning node. A new method of exploration is also proposed to increase the learning performance. Two basic robot behaviours which are a goal-seeking and an obstacle avoidance behaviour are simulated to show the promise of the proposed techniques. The goal-seeking behaviour is implemented on a real robot. The experimental results show that this method is practical for a real-world problem.

Stable adaptive bilateral control of transparent teleoperation through time-varying delay

Passivity concept has been using as a framework to solve the stability problem in bilateral control of telemanipulation. However, the conservative selection of dissipating element applied to maintain system stability in network communication leads the system to imperfect operation or losing transparency. In this paper, we proposed a new control scheme to adapt characteristic impedance with time. The proposed method is not only presented in simple form, but also effectively make the operation transparent teleoperation. We verified the validity of our method by teleoperation simulations with constant and variable time delay.

Telerobotic remote handling of protein crystals

A combined university/industry team has developed a prototype system for handling protein crystals aboard the space station. This system uses a miniature direct drive robot, CCD television cameras, and a client-server computing system using Internet protocols to support the capture of protein crystals from aqueous growth solutions. The system was demonstrated between Huntsville AL. and Seattle WA. An operator in Huntsville controlled the mini robot by invoking predefined relative and absolute macro files. The operators observed results using video images sent through the Internet link using Cu-SeeMe video conferencing software. In 3 of 4 trials, the operators successfully captured 0.5 mm simulated protein crystals into a glass capillary. The system is a promising start for the development of a space-station based remote protein crystal analysis facility.

Haptic feedback of kinematic conditioning for telerobotic applications

Kinematic conditioning of robot manipulators is the problem where small motions in Cartesian space cause excessive joint velocities. This problem is significant in teleoperation. Haptic feedback provides the bi-directional flow of information which allows the operator to control the telerobot interactively. Haptic feedback of kinematic conditioning is proposed as a new approach to achieve better performance in telerobotic control near kinematic singularities. Four different singularity force feedback methods are defined and studied. Experimental results with a force feedback master and simulated slave system show that teleoperation performance near singular configurations was affected and improved by using singularity force feedback.

Augmented haptics of manipulator kinematic condition

This paper describes a study of whether haptic feedback can be used to represent information that is normally difficult to obtain via visual feedback in telerobotic system. Problems of manipulator kinematic condition such as singularity and joint limit have been well known for a long time. Kinematic condition of the manipulator is difficult to be recognized visually. Poor kinematic condition often causes trajectory error or other undesirable effects in the system. This problem is quite significant in telerobotics since a fully pre-planned path that completely excludes poor kinematic condition is usually not available.

Motion classification using IMU for human-robot interaction

This paper proposed the method for identifying motion of the robot using signals from an Inertial Measurement Unit (IMU). When human interacted with the robot, the motion of the robot needed to be classified in order to create the suitable emotions of the robot during the human-robot interaction process. The output data from the IMU were classified by applying heuristic conditions to the high and low frequency data. Wavelet transform was added in the pre-processing step for finding the starting point of the motion state. We found that most of our tested actions could be correctly classified using our proposed method with 100% accuracy.

Design of pipe crawling gaits for a snake robot

This research aims to study the parameters that contribute to the crawling performance of the snake robot inside an inclined pipe. The shape and motion propagation of the snake robot directly affect the forward crawling speed of the robot inside a pipe. Motion shape code is proposed as the simplified method for describing shape of a modular snake robot. The motion propagation can be created by shifting the code to the adjacent joint. Five different motion shapes were tested on the 12-joints modular snake robot that moved inside a pipe with varied inclined angles. The experimental results showed that the motion shape that has the maximum number of waves along the robot length and the largest basis angle resulted in the fastest forward crawling speed.

Mechanics of cooperative nonprehensile pulling by multiple robots

The objective of this paper is to analyze the mechanics of object pulling by more than one robot using flexible tools such as rope. The concept of nonprehensile pulling where an object is not completely tied by a rope such that an object can roll or slide along the rope is the main focus of this study. The necessary condition for a stable nonprehensile pulling is defined. The mechanics of nonprehensile pulling is used to define the set of feasible motion directions of the object. This set of feasible motion directions of the object is then used for planning the coordination between two robots such that the task of object transportation can be achieved. The proposed analysis can also be extended for multiple objects transportation. This study provides a necessary basis required in developing an algorithm for a group of robots to autonomously cooperate such that the pulled object can be positioned and oriented in the 2D space.

Impact reduction mobile robot and the design of the compliant legs

In most mobile robots, the ability to move from point to point in various types of terrain was the most crucial part to the design. Being able to survive through impact conditions is also essential for robots under hazardous circumstances such as rescue robots or military robots. In this paper, we designed and developed a robot with impact reduction mechanism which is based on the compliant design of its legs. The stiffness of the legs was designed to not only to serve walking purposes but also to help reduce the impact while dropping. An experiment was set to investigate how the radius of curvature of the connecting plate and the compliant leg of the robot play a role in impact absorption. The radius of curvature is one of the key factors which vary the stiffness of the compliant parts. With this design, the robot will gradually press the ground during landing using springlike legs. The compliant legs with nonlinear spring constant help absorb impact energy while the robot hits the ground. During drop-landing motion, the robot also transforms itself from a spherical shape into a legged robot while landing. The legs are extended into a walking mechanism on uneven terrain and retracted to create a ball shaped robot for rolling motion over smooth terrain. The transformation between the spherical shaped robot and the legged robot increase its motion capabilities under various conditions including falling, rolling and walking.

Comparison and analysis of expertness measure in knowledge sharing among robots

Robot expertness measures are used to improve learning performance of knowledge sharing techniques. In this paper, several fuzzy Q-learning methods for knowledge sharing i.e. Shared Memory, Weighted Strategy Sharing (WSS) and Adaptive Weighted Strategy Sharing (AdpWSS) are studied. A new measure of expertise based on regret evaluation is proposed. Regret measure takes uncertainty bounds of two best actions, i.e. the greedy action and the second best action into account. Knowledge sharing simulations and experiments on real robots were performed to compare the effectiveness of the three expertness measures i.e. Gradient (G), Average Move (AM) and our proposed measure. The proposed measure exhibited the best performance among the three measures. Moreover, our measure that is applied to the AdpWSS does not require the predefined setting of cooperative time, thus it is more practical to be implemented in real-world problems.