A collision free motion of a humanoid robot is perturbed when a balance controler is applied to make the robot stable, then the motion may cause collisions. This paper presents a collision avoidance algorithm of humanoid robots that can be used together with a balance controller. The proposed algorithm cancells the perturbations at shoulder joints, and the sweeping volume of an arm can be kept same even when a balance controller is applied to the robot. The prerequisite of the algorithm is that three axes of shoulder joints should pass through a point.
This paper discusses the manipulation of a big and heavy object by a human type robotic mechanisms. When a human type robot pushes an object, slip may occur at the contact point between the hand and the object and at the contact betweem the leg and the floor. To avoid the slip, we obtain the region of the internal force acting among the contact points. We also obtain the region of the foot position enabling a robot to push an object effectively. To verify the effectiveness of our method, we show simulation results.
This study aims at the construction of the control system to have a robot approach to an object and move to an objective point. The designed autonomous mobile robots are equipped with an omni-directional camera for a sensory system. These cameras provide precise directional information, but suffer from much noise on distance, since there are visual noises in dynamical environment. To cope with these noises the fuzzy control theory is adopted. The rules for going to an object, wraparound and obstacle avoidance are represented by Potential Membership Functions (PMF) , and the rules are merged by fuzzy operation, so that an agreement point of each rule can be found well. The characteristics of PMF are easy to understand intuitively, since the horizontal axis is based on direction and the grade represents the priority of the direction. Additionally because PMF includes the potential factor, it is possible to use the grade for the desired strength against the direction. To apply this grade for the speed element, the robot is able to decrease the speed naturally near the obstacles. The usefulness of this control system is investigated theoretically and experimentally.
This paper proposes a new mobility system for planetary exploration robot using springs and linear actuators. In microgravity environment, it is difficult to obtain the horizontal velocity, because the friction force between a robot and the ground is very small. By pushing the ground, however, the robot can not only hop, but also obtain the horizontal velocity. The proposed hopping robot consists of two masses, and can push the ground by using springs. The robot can hop from the stationary state by transforming the elastic energy to the kinetic energy using the sprigs and linear actuators. Furthermore the robot can land without bounding by transforming the kinetic energy to the elastic energy. The simulation studies and the ground experiments show the effectiveness of the proposed mobility system.
This paper gives the advanced control of automatic pouring process, with special attention paid to the realization of the expert's skill in pouring process. First, in order to realize the expert's skill, the novel Automatic Pouring Robot (APR) is developed and the dynamics of the pouring process is clarified. Second, the pouring model is constructed. In addition to the sloshing control by Hybrid shape approach and multi-axes synchronized control, the APR is controlled by the supervisory control system that reasonably swiches the controllers designed by H∞ control theory to control the liquid level in a sprue cup and model predictive control to control the outflow from the ladle. Finally, the effectiveness of the proposed system is shown through simulations and experiments.
The authors proposed a new concept for the integration of locomotion and manipulation, “the Integrated Limb Mechanism”. The concept covers the technical fields both of the control integration and of the mechanism integration necessary for the practical working robot. This paper describes the mechanical configuration design of a prototype hexapod with integrated limb mechanism, named “MELMANTIS-1”. MELMANTIS-1 uses a six-bar linkage with four degrees-of-freedom as an example of integrated limb mechanism with both advantages for leg and for arm. Several items on leg and body configuration design are taken into account for a hexapod with the transformable legs into arms. A kinematic analysis on the volume of dual or triple arm workspace is performed for the evaluation of the mechanism and configuration. Basic motion of MELMANTIS-1 is experimented to verify the transformation of a leg mechanism into an arm.
This paper describes a map generation method using an omnidirectional stereo and a laser range finder. Omnidirectional stereo has an advantage of 3D range acquisition, while it may suffer from a low reliability and accuracy in range data. Laser range finders have advantage of reliable acquisition of data, while they usually obtain only 2D range information. By integrating these two sensors, a reliable map can be generated. Since the two sensors may detect different parts of an object, a separate probabilistic grid map is first generated by temporal integration of data from each sensor. The resultant two maps are then integrated using a logical integration rule. An ego-motion estimation method is also described, which is necessary for integration of sensor data obtained at different positions. Experimental results on autonomous navigation in unknown environments show the feasibility of the method.