Basically Biped robots had to walk like human, to generate this system we use basic approach for analysis human walking behavior. Human walking behavior can be analyzed using Zero Moment Point (ZMP). Zero Moment Point (ZMP) is a zero point which total of gravity forces and horizontal inertia equal 0 (Zero). This paper describes about early step to made basic walking trajectory for FLoW robot. To determine the best Zero Moment Point (ZMP) reference, it can use the comparison of distance and time. From this comparison, the walking characteristic of FLoW robot was found and the best Zero Moment Point (ZMP) trajectory also founded. The basic walking trajectory will generate based on the best Zero Moment Point (ZMP) trajectory as a reference. In the FLoW robot, the best Zero Moment Point Trajectory was founded at distance of 14cm with difference of Zero Moment Point Reference and Real Zero Moment Point equal 0.012cm.

Unicycle mobile robot is wheeled mobile robot that can stand and move around using one wheel. It has attached a lot of researchers to conduct studies about the system, particularly in the design of the system mechanisms and the control strategies. Unlike two wheel balancing mobile robot which mechanically stable on one side, unicycle mobile robot requires additional mechanisms to keep balancing robot on all sides. By assuming that both roll dynamics and pitch dynamics are decoupled, so the balancing mechanisms can be designed separately. The reaction wheel is used for obtaining balancing on the roll angle by rotating the disc to generate momentum. While the wheeled robot is used for obtaining balancing on the pitch angle by rotating wheel to move forward or backward. A PID controller is used as balancing control which will control the rotation motor on the reaction disc and wheel based on the pitch and roll feedback from the sensor. By adding the speed controller to the pitch control, the system will compensate automatically for perfectly center of gravity on the robot. Finally, the unicycle robot will be able to balance on pitch angle and roll angle. Based on simulation result validates that robot can balance using PID controller, while based on balancing pitch experiment result, robot can achieve balancing with maximum inclination about ±23 degree on pitch angle and ±3.5 degree on roll angle with steady state error 0.1 degree.

Pens-Wheel is an electric vehicle which uses one wheel that able to balance itself so the rider does not fall forward or backward while riding it. This vehicle uses one brushless DC motor as actuator which capable to rotate in both directions symmetrically. The vehicle uses a combination of accelerometer and gyroscope contained in IMU (Inertial Measurement Unit) for balancing sensor. The controlled motion on the vehicle occurs only on the x-axis (pitch angle), in forward and backward directions. The PID (Proportional Integral Derivative) control algorithm is used to maintain the balance and movement of the vehicle. From the simulation and application in the real vehicle, the use of PID control is capable driving the vehicle in maintaining the balance condition within ±10° tilt angle boundary on flat surface, bumpy road, and inclining road up to 15° slope.

Many electrical vehicles have been developed recently, and one of them is the vehicle type with the self-balancing capability. Portability also one of issue related to the development of electric vehicles. This paper presents one wheeled self-balancing electric vehicle namely PENS-Wheel. Since it only consists of one motor as its actuator, it becomes more portable than any other self-balancing vehicle types. This paper discusses on the implementation of Kalman filter for filtering the tilt sensor used by the self-balancing controller, mechanical design, and fabrication of the vehicle. The vehicle is designed based on the principle of the inverted pendulum by utilizing motor's torque on the wheel to maintain its upright position. The sensor system uses IMU which combine accelerometer and gyroscope data to get the accurate pitch angle of the vehicle. The paper presents the effects of Kalman filter parameters including noise variance of the accelerometer, noise variance of the gyroscope, and the measurement noise to the response of the sensor output. Finally, we present the result of the proposed filter and compare it with proprietary filter algorithm from InvenSense, Inc. running on Digital Motion Processor (DMP) inside the MPU6050 chip. The result of the filter algorithm implemented in the vehicle shows that it is capable in delivering comparable performance with the proprietary one.

The Inverted pendulum platform is an example of classic unstable control system. Even though the system has been fairly tested and documented, it still draws attention of many researchers due to its application in unicycle robot. In the unicycle robot, there are problems that arise control strategy in the reading position of the robot tilt. This paper proposes to use the Kalman Filter Estimation angle for data processing Inertial Measurement Unit (IMU) to obtain estimates of the robot tilt position. In the previous study also found problems when using only one relatively low speed IMU sensor obstacles that the response given by the sensor. This paper uses two IMU sensor readings to speedup the response of the sensor and get accurate data during a shorter period. The proposed algorithm uses a new sensor placement strategy on a rigid body robot, with a reading sensor in interleaved mode. Kalman Filter algorithm incorporating placement constraints to achieve the estimated position of the robot tilt angle accurately. The results show synchronization time sampling of the two Inertial Measurement Unit (IMU) sensor improves the response and a twice faster in estimating the position of the robot tilt compared to the use of one sensor. Merging time sampling 2 sensors can be applied on a unicycle robot in order to have a quick response to the reading of the tilt position of the robot.

This paper described about humanoid robot dancing which designed like human and can dance like the real dancer with flexibility of human dancer. Part of link and frame of this humanoid robot dancing was built from 3D printing technique. Dance movement will be resemble the motion system possessed by humans which design of robot seize on mechanical approach, and will be controlled by a motion program to be included in the motion system. Humanoid currently developed to resemble human beings, including the shape of human skeleton and their functions. During this period, humanoid robot dancing rare develop a complex system of abdominal motion system. Therefore, withdrawn hypothesis presence of abdominal motion system can improve their existing motion systems. Some possible methods used in these systems, such as inverse kinematic, and forward kinematic. Trajectory dancing style rise from two kinds of kinematic. Inverse kinematic will controlled any motion of leg motion system. This system will hand over degree value in joint space as output. Position in Cartesian coordinate space as input. Timing control for every step from primitive style is needed. Therefore, actuator of ankle will not burden in execution of working because it will hand over some smoothness in every single last step in execution. Forward kinematic controlled abdominal motion system and give Cartesian coordinate space as output. From that method we were successfully built trajectory dancing modelling of humanoid robot dancing.

This paper presents neuro-based push recovery controller applied in humanoid biped robot in order to keep the stability with minimum energy required. There are three motion patterns in human behavior when it gets external perturbation, those are ankle behavior, hip behavior, and step behavior. We propose the new model of modular recurrent neural network (MRNN) for performing online learning system in each motion behavior. MRNN consists of several recurrent neural networks (RNNs) working alternately depending on the condition. MRNN performs online learning process of each motion behavior controller independently. The aim of push recovery controller is to manage the motion behavior controller by minimizing the energy required for responding to the external perturbation. This controller selects the appropriate motion behavior and adjusts the gain that represent the influence of the motion behavior to certain push disturbance based on behavior graphs which is generated by adaptive regression spline. We applied the proposed controller to the humanoid robot that has small footprint in open dynamic engines (ODE). Experimental result shows the effectiveness of the push controller stabilizing the external perturbation with minimum energy required.

The development of microkernel has sharply increased. One of the most successful microkernel implementation is L4. L4Linux is L4 version that is able to run virtualized Linux. We have also built our microkernel named FLoW. In this paper we described about our achievement in developing virtualized Linux on top of our FLoW microkernel. We implemented unique design about virtualizing more than one Linux. The idea is about creating more than one Linux to handle the requests made by user application about Linux system call. So every user application that requests the feature of Linux system call will be handled by more than one Linux Service. The result shows that there is an acceleration of processing time. The total CPU cycle used by user application in working with user request are less when having such mechanism. That means our mechanism successfully accelerate around 17% processing time of user application that request Linux system call from within our Linux Service that run on top of FLoW Microkernel. Then the system endurance testing resulted that our system is able to handle around 96 Linux system calls request almost at the same time without break.

One of the focus in humanoid robot research is head motion and mechanism, yet it still has a limited pan and tilt motion to represent a real human head movement. This study is working on kinematic analysis of an earlier proposed 7 DOF manipulator system that can closely mimics a human head motion capability. The concentration of the work is to design a head mechanism and to develop a composite kinematics model of the head motions which includes the eyes, the head itself and the neck. Such a model was also simulated using the dynamic V-REP application to verify its movements capabilities and to compare the kinematics analysis. Stages will be discussed in more detail in the chapter on testing. The results showed transformation biggest difference is 2.3340% error that occurred on 3 DOF motion contained in the head manipulator.

Robotics has been widely used in education as a learning tool to attract and motivate students in performing laboratory experiments within the context of mechatronics, electronics, microcomputer, and control. In this paper we propose an implementation of cascaded PID control algorithm for line follower balancing robot. The algorithm is implemented on ADROIT V1 education robot kits. The robot should be able to follow the trajectory given by the circular guideline while maintaining its balance condition. The controller also designed to control the speed of robot movement while tracking the line. To obtain this purpose, there are three controllers that is used in the same time; balancing controller, speed controller and the line following controller. Those three controllers are cascaded to control the movement of the robot that uses two motors as its actuator. From the experiment, the proposed cascaded PID controller shows an acceptable performance for the robot to maintain its balance position while following the circular line with the given speed setpoint.

In this paper we described a model and a simulation of walking pattern of FLoW bipedal robot. This kinematics design combined four-bar linkages and translational actuators. Inverse kinematics problem is solved by using trigonometry approach. While walking pattern is made of a superposition between linear and sinusoidal function. The equations is simple and does not required long computation time. So it is efficient to maximize overall process on robot. The stability of the robot is controlled by CoM point which is calculated by multiplying the mass and position of the robot element and keeping it remain in the robot support polygon.

The development of technology for supporting learning system at this time takes place very rapidly. Human Computer Interaction provides users that allows to control presentations in a natural way by their body gestures and voice commands. The authors propose a simple system that can be use to control presentation by using hand gestures and also by using Indonesian speech recognition. Our proposed system consist of gesture recognition with Kinect 3D skeletal data, and utilize Google Speech API to recognize Indonesian language as voice command. In this system we built applications to perform hand gestures and speech recognitions and utilize libraries to perform mouse and keyboard actions in presentation application. From the experiments, hand gestures application can process in realtime with the accuracy more than 90%, and for Indonesian language speech recognition gave accuracy almost 99% with average speed of recognition around 1 second.

In this paper, we described a model and a simulation of forward and inverse kinematics of a parallel link leg performing a predefined hula-hoop motion of FLoW, a bipedal humanoid robot. This lower body leg having 12 DOF links and joins configuration were described using D-H parameters. It was assumed that the motion is slow enough. Thus, by keeping the centre of mass projection to the floor to be always inside the support polygon of the robot the system can be regarded as stable.

Balancing robot is a robot that relies on two wheels in the process of movement. Basically, to be able to remain standing balanced, the control requires an angle value to be used as tilt set-point. That angle value is a balance point of the robot itself which is the robot's center of gravity. Generally, to find the correct balance point, requires manual measurement or through trial and error, depends on the robot's mechanical design. However, when the robot is at balance state and its balance point changes because of the mechanical moving parts or bringing a payload, the robot will move towards the heaviest side and then fall. In this research, a cascade PID control system is developed for balancing robot to keep it balanced without changing the set-point even if the balance point changes. Two parameter is used as feedback for error variable, angle and distance error. When the robot is about to fall, distance taken from the starting position will be calculated and used to correct angle error so that the robot will still balance without changing the set-point but manipulating the control's error value. Based on the research that has been done, payload that can be brought by the robot is up to 350 grams.

Research on the control of movement omnidirectional on mobile robots have been much done. On the problems there are in the Asia-Pacific Robot Contest (ABU Robocon) 2014, about the trajectory of the line on a robot that cannot be used as a reference point for the trajectory of the movement from the robot. If using the line as a reference and no field the entire area cannot be mapped, so that only part that there is a line of course that can be passed by a robot. Some method of tracking the path that is used will form new path when the robot out of the set so that the movement of the robot not appropriate. Many researchers has done research on a method of tracking the path on omnidirectional use to maintain control of robots to maintain in order to stay on track. Pure pursuit and bicycle path planning are other methods used path tracking. This paper the researchers did research on comparing the two method, there are pure pursuit path tracking and bicylce path planning. From both of these methods, methods bicycle path planning RMSE values to track the movement of the robot is smaller than the pure pursuit method. With the smallest RMSE value of the bicycle path planning methods at 2.59 cm and pure pursuit method 3.10 cm, while the largest of the methods bicycle path planning methods 7.9 cm and 7.57 cm pure pursuit.