The four-wheeled omnidirectional platform is great to use for an indoor mobile robot. It can increasingly move and change heading directions. However, the robot is easy to slip when it is moving. One or more wheels are sometimes not touching the ground. This paper approaches to solving these problems by computational simulation in locomotion. The mathematical models simulate the robot movement. The robot struggles to go to the target with randomly simulated slips and untouched ground circumstances. The robot can estimate the positions using the Kalman filter and readjust itself to the planned path. Consequently, this paper demonstrates the motion improvement and compares the results of decreasing errors.

The conventional wheel is famous in the industrial mobile robot because it is simple to build and easy to control and to maneuver. On the other hand, the legged robot is complicated to control but has a high performance in locomotion like highly evolved legged-creatures. This research describes the four-legged robot design platform with the soft feet, which are made by siliconematerial forming. The robot's feet are implanted by the strain gauge sensor. The robot can percept the external force sensitively when its feet touch something. The research demonstrates the forward kinematics and inverse kinematics of a four-legged robot with three joints for each. The result has validated the kinematic equations by implementing them on the real dog-like pet robot. The robot walked with a pure walking gait and showed a sensing signal from the soft foot and ground contact.

No abstract

No abstract

In order to improve the intelligent service robot, the visual perception is a crucial. This paper presents the human detection for a service robot. The camera is installed at the robot by uprisen angle around 30 degree. Two feasible algorithms for the real-time detection between Haar cascade and Single Shot Detector (SSD) algorithm are compared. This research collects the training data as 1,000 images by the uprisen angle and the different views of the human, within 60 cm of user range. between a camera and human. The result shows that our proposed model of SSD method is higher performance than another by 0.933 of the average of IoU. Therefore, the proposed method is suitable to apply for the service robot.