This paper presents the development of the ring sensor to monitor a patient 24 hours a day for a tele-nursing system. The ring sensor is worn by the patient at all times, hence the health status is monitored 24 hours a day. The sensors packed into the ring include LEDs with different wavelengths, and technologies of photoplethysmography and pulse oximetry are implemented on the ring. The sensor data are transmitted to a computer through the digital wireless communication link and the patient status is analyzed continually and remotely. Any trait of abnormal health status and possible accidents is detected by analyzing the sensor data. A combination of a global receiver and multiple local ones are used to estimate the patient's location and activity. Both the physiological data and the position information can be used to make an accurate decision as to whether a warning signal must be sent to a medical professional caring the patient. An issue of power reduction for miniaturization of the ring sensor is also addressed.

In this paper, we present a new approach to learning robot impedance control parameters from a logic task description. In this approach, we first describe the desired behaviour of a robot for performing a given task at a logic level. A simple logic branch control using a quasi-static force-to-motion map is created based on the logic description. The progressive learning method is then applied to this logic branch control in order to create a dynamic control, i.e. impedance control, for performing the task quickly and dynamically. Starting with a simple logic description about the robot behaviour, the system can develop a fully dynamic impedance control by progressively learning the process dynamics. The problem is formulated in the context of high-speed insertion, and the proposed approach is verified through simulation.

A novel approach to the design of trajectories for automatic tuning of non-collocated sensor/actuator systems is presented. This approach uses the progressive learning theory, a method for tuning many control parameters in a particular sequence so that all the parameters can be tuned stably and effectively. Using this progressive learning method a sequence of training trajectories can be designed, which not only satisfy geometric constraints for performing a task but also take into account the dynamics of the machine to better perform the task. This approach is implemented on a single degree-of-freedom high speed linear slider, and experimental results and discussions are presented at the end.

An approach to learning control using an excitation scheduling technique is developed and applied to an impedance learning problem for fast robotic assembly. Traditional adaptive and learning controls incur instability depending on the reference inputs provided to the system. This technique avoids instability by progressively increasing the level of system excitation. Called progressive learning, it uses scheduled excitation inputs that allow the system to learn quasistatic parameters associated with slow input commands first, followed by the learning of dynamic parameters excited by fast input commands. As learning progresses, the system is exposed to a broader range of input excitation, which nonetheless does not incur instability and unwanted erratic responses. In robotic assembly, learning starts with a slow, quasistatic motion and goes to a fast, dynamic motion. During this process, the stiffness terms involved in the impedance controller are learned first, then the damping terms and finally by the inertial terms. The impedance learning problem is formulated as a model-based, gradient following reinforcement learning. The method allows the suppression of excessive parameter changes and thereby stabilizes learning. By gradually increasing the motion speed command, the internal model as well as the control parameters can be learned effectively within a focused, local area in the large parameter space, which is then gradually expanded as speed increases. Several strategies for motion speed scheduling are also addressed.

A novel approach to stable learning control is developed inspired by human learning behavior, and applied to an impedance learning problem for high-speed dynamic robotic assembly. The new method termed "progressive learning" uses scheduled excitation inputs that allow the system to learn quasi-static, slow modes in the beginning, followed by the learning of faster modes. This new method is presented in the context of high speed robotic assembly, where an impedance control law is learned with this excitation scheduling method. Extensive simulation results are provided to demonstrate the effectiveness of this method. A detailed analysis of the mechanism of progressive learning is also provided and verified through simulation.

The authors develop a method of modeling a human manipulative skill using human linguistic knowledge about the task. A global nonlinear structure of human control behavior is constructed based on the linguistic information, and all functionalities used by the linguistic structure are identified from human demonstration data. Mapping between sensor space and human mental space for input signals is discussed to elucidate human skills. Techniques for selecting significant features extracted from sensor signals and reducing the dimension of the sensor space are developed. The techniques were applied to a direct-drive deburring robot to verify the feasibility of the method.>

An approach to the teaching of manipulative skills is developed and applied to a deburring robot. Teaching data acquired from a human expert, who can perform an efficient job, are processed on a computer to attain his skilful manipulation strategies. The strategies are described by a group of control laws that relate sensor signals to motion commands. Sensor signals are processed by using pattern recognition techniques to interpret sensor information and to allow the robot to recognize the state of the process. The control laws designate which control action the robot should take in response to each signal pattern generated in the process. A method for driving a compact set of discrimination functions for real-time recognition of signal patterns is also discussed. The method is implemented on a direct-drive deburring robot. It is demonstrated that the robot can mimic the skilful manipulation of the human expert and perform the task efficiently.>