The flexible transfer system (FTS) is a self-organizing manufacturing system composed of autonomous robotic modules, which transfer a palette carrying machining parts. Where, the central issue is realization of both of higher efficiency and flexibility to cope with environmental change, such as a sudden change of machining plan or break-downs of the modules. Through the self-organization of a multi-layered strategic vector field corresponding to a task, the FTS can generate quasi-optimal transfer path with Learning Automata. Also the optimal planning is attempted by use of Genetic Algorithms, which bases on the global information on the system. In this paper, we propose a hybridization method between the distributed and centralized approaches. Simulation conducted to evaluate the basic system performance and the results show the effectiveness.
This paper describes a novel concept of driver assistance system that can give timely advice on safe and efficient driving. From both the current traffic condition obtained from sensory data and the driver's goal and preference in driving, it autonomously generates advice and gives it to the driver. Not only operational level advice, such as emergency braking due to an abrupt deceleration of the front vehicle, but also tactical level advice, such as lane changing due to the congested situation ahead, can be generated. Two main components of the intelligent navigator, the advice generation system and the road scene recognition system, are explained. On-line experiments using the prototype system show the potential feasibility of the proposed concept.
Robot and virtual reality technologies may be applied to rehabilitation training so that fewer human therapists are needed, training can be quantitatively evaluated, and training can be more enjoyable to patients. A training system in rehabilitation can use robot or other virtual reality technologies, but it must be far safer than ordinary industrial robot systems. The authors have been studying actuators, which are safe and easy to control, and have developed force display systems using such actuators. As an application of force display technology, they have developed a rehabilitation training system using ER actuators. The authors have also developed a computer program for the system, which evaluates the physical capability of the upper limbs of patients and assists in their rehabilitation training. This paper is a summary of the basic experiments with the system as applied at a hospital to patients suffering from upper limb paralysis.
This paper describes a concurrent logic programming language MRL for use in developing programs to cooper-atively control multiple robots. MRL describes actions of each robot and sensor as sets of logical formula. MRL enable us to easily implement complex tasks such as concurrency control, cooperation and negotiation between pro-cesses, and emergent event handling for multiple robots. We conducted an experiment on program development for paper delivery task by cooperation with mobile robots, manipulators and cameras to demonstrate the advantages of MRL programming framework. The results indicated that the MRL programs were more abstract and natural than conventional procedure-oriented programs, resulting in realization of flexible cooperation. Since MRL programs are compiled into C programs with little overhead, NIRL is useful as a multiple robot programming language efficient in both program execution and development.
We have proposed the sway compensation trajectory of robot body to realize dynamically stable walking for a quadruped walking robot. This method uses the lateral motion of the robot body to keep a zero momentum point (ZMP) on a diagonal line between the support legs. However, energy consumption of the sway compensation trajectory is larger than the case without trajectory control because periodical acceleration of the body has to be generated to realize sway motion. In this paper, we propose a new sway compensation trajectory for a quadruped walking robot named 3D sway compensation trajectory. This method is the extension of conventional sway compensation trajectory toward three dimensional motion, and the position of ZMP is controlled not only by lateral body motion but also by longitudinal and vertical body motion. Next, we demonstrate that energy consumption for walking can be reduced by combining these motions through theoretical analysis and computer simulation, and an optimum 3D sway compensation trajectory that minimizes energy consumption is proposed. Optimum trajectory is applied to a quadruped walking robot and the validity of the proposed trajectory is verified using the dynamic motion simulator, ADAMS.
A newly designed four wheeled omni-directional vehicle which can vary its footprint is presented. Each wheel (made with a normal rubber tire) is attached with an offset to the tip of a link through a rotary joint and the another tip of the link is installed to the robot body with a rotary joint. The former joint is called “elbow” and the later one is “shoulder.”Individual joint has a DC servo motor and an electro-magnetic brake to control its angle. Thus, this mechanism consists of 12 motors and 8 brakes. First the concept of the proposed mechanism is described, and its kinematic behavior is analyzed. The principal part of the analysis is to derive a continuity condition of the elbow angle to design the trajectories for non-side-slip motions. The analysis is verified by some experimental results using a prototype one. Finally, the effectiveness of the proposed mechanism to gain the running stability based on controlling the zero moment point (ZMP) is also discussed.
In this paper, we describe a control method for a multi-fingered hand with uncalibrated vision to manipulate an object. To achieve a robust and quick manipulation without precise calibration, adaptive visual servoing is applied. on the other hand, internal force control also has to be applied to maintain grasping. To integrate these controllers, we utilize the concepts of visual impedance and the principle of virtual work, and eliminate the interference. Experimental results are shown to demonstrate the validity of the proposed scheme.
In this paper, we introduce a “hybrid”controller for better regulation of the swing up behavior of a two degree of freedom brachiating robot. In this controller, a previous “target dynamics” controller and a mechanical energy regulator are combined in a suitable fashion. The proposed controller guarantees the boundedness of the total energy of the system. Moreover, numerical simulations as well as experiments suggest that this hybrid controller achieves much better regulation of the desired swing motion than the target dynamics method by itself.
This paper describes mobile manipulation of humanoid robots. A humanoid robot is a kind of integrated machines: a two-arm and two-leg mechanism. It could be well applied to mobile manipulation in non-routine task automation. The main objective of the mobile manipulation is to obtain versatile and efficient manipulation by the arms. In this respect the legs are required to assist the arms obtaining their high manipulability by changing step length and tim-ing of step motion. The required motion of the legs in mobile manipulation is different from the conventional biped locomotion. From this point of view, we discuss and analyze the required motion of legs in mobile manipulation. We propose a basic idea to quantitatively evaluate robot performance in mobile manipulation. Because manipulability and stability vary during the robot motion, we take an average and a variance of manipulability measure of arm, and an average and a minimum of stability margin per a locomotion cycle as an evaluation criterion. By computer simulations, the relationship between these evaluations and some patterns of leg motion is analyzed by using the proposed control method of mobile manipulation. The simulation results show that the leg motion can vary the manipulability of the arm by changing the step length and the timing of the step motion. It is also found that the manipulability of the arm and the stability of the robot are closely related.
In human robot behavioral interaction, characteristic of the robot's reaction-decision algorithm produces impres-sions that the human user has about the robot. The purpose of this paper is to analyze the relationship between robot's interactive characteristic and produced impression about the robot in a human-robot-interaction experiment. The analysis was done quantitatively by using methods of information theory. The robot's interactive characteristic is quantified as information transfer efficiency of Shannon information theory; the algorithm that selects a fixed certain reaction in respect to certain input (1-to-1 reaction) has very high informational transfer efficiency. In con-trast, very low informational transfer efficiency of the interaction means the robot makes reaction almost randomly. In the experiment, 3 reaction algorithms of the robot were prepared; the informational transfer efficiencies of them are 1.0, 0.5, and 0.0. The result shows that high transfer efficiency produces impressions that the robot has strong intention. Also it is shown that maximum of degree of interest on interaction was got at moderate transfer efficiency (0.5) . Those results will be utilized in interaction algorithm of robots that adjusts robots' behavioral characteristic in order to produce proper impression toward humans.