Son Kuswadi

The Study on Optimal Gait for Five-Legged Robot with Reinforcement Learning

The research of legged robot was rapidly developed. It can be seen from recent ideas about new systems of robot movement that take ideas from nature, called biology inspired. This type of robot begins replacing wheeled robot with various functions and interesting maneuvers ability. However, designers should decide how many legs are required to realize the ideas. One of the ideas that are rarely developed is odd number of legs. This research focused on five legs robot that inspired from starfish. To realize the intelligent system in robot that does not depend on the model, this research used reinforcement learning algorithm to find the optimal gait when robot is walking. In order to achieve this goal, trial and error have been used to provide learning through an interaction between robot and environment based on a policy of reward and punishment. The algorithm is successfully implemented to get the optimal gait on a five-legged robot.

Adaptive morphology-based design of multi-locomotion flying and crawling robot “PENS-FlyCrawl”

Usually, robot is designed by using only one locomotion mechanism, such as walking, hopping, and jumping. The reason, among others, are simplicity, saving energy, and focus on single task and mission. This paper deal with design and implementation of multimodal flying and crawling robot by using adaptive morphology. The aim of this development is to provide robot platform that could achieve multiple locomotion mechanism in order to perform flexible and challenging mission. This kind robot is required in the field like collapse building due to earthquake that cluttered, unstructured and unstable.

An integrated project-based learning for Intelligent Control lecture and practice at polytechnics graduate school

This paper presents an integrated project-based learning of lecture and laboratory practice on Intelligent Control course in Master degree Program, at Politeknik Elektronika Negeri Surabaya (PENS). This first skill-based graduate engineering education program in Indonesia, was started since 2013 to fulfill the lack of professional engineers in advanced level to be ready as innovators and leader in industries. Moreover, since the role of PENS as National Resources Polytechnics in electrical and information related fields, this program also provide practical-oriented graduate level education for the countrys polytechnic lecturer that inline with the nature of polytechnic education. We will present one of example how we deliver our integrated knowledge and skill to the students, in this case Intelligent Control course. The course enable student learn the knowledge given by lecturer in the classroom, and - together with lecturer - the students design and implement practice in laboratory to get hands on practical ability. By project-based learning, students more actively involve in the learning process, and hence the student-centered learning could be realized.

Temperature distribution pattern in normal and cancer tissues: the effect of electromagnetic wave

Abstract : Temperature distribution pattern on normal and cancer tissues caused by the Electromagnetic wave effect is presented in this paper. First, the pattern is modelled by using Bio Heat Transfer Equation and then the Galerkin Finite Element method is used to solve the equation via computer simulation. From the simulation results it can be shown that the temperature change on the cancer tissues is relatively constant and 0.4 degrees Celsius higher than the change on the normal tissues. From this phenomenon it can be understood that the damage of the cell of the cancer tissues is greater than the damage of the cell of the normal tissues. Therefore it is possible to use the Hyperthermia technique with electromagnetic wave for cancer therapy. The simulation results can be also used to decide the control strategy for giving the dose energy of electromagnetic wave which is flowed out to the cancer tissues.

Effective Intelligent Control teaching environment using Challenge Based Learning

This paper deals with development of effective Intelligent Control teaching environment using Challenge Based Learning (CBL) at Mechatronics Department of Politeknik Elektronika Negeri Surabaya (PENS). CBL is one of problem-based, student centered learning in which a group of students learn according their own pace, and they make their own final challenging target. Moreover, social media was utilized to communicate with their peers around the world. The CBL was implemented in one pre-selected class, and another class was teach by using conventional teacher-based learning, as control variable. Both classes then evaluated by using the same problem in mid and final examination. Satisfaction of effective teaching environment then verified by using questionnaire. It was revealed that the students consider that the CBL method quite enjoyable and attractive learning environment.

Proposal of a Warning Receiver System based on open pervasive display network

Indonesia is prone against local tsunami. After Aceh Tsunami in 2004, many works have been done to reduce disaster risk. One of them is InaTEWS, which is an end-to-end tsunami early warning system. Currently, InaTEWS support 5 types of warning receiver system (WRS): siren, SMS, telephone, fax, and website. In here, we propose a design of WRS based on pervasive display vision. Where it uses a networked public display to build a distribution platform. It can be used to distribute content and application. Here we describe a system design and some use cases of the platform in the context of a WRS and as a medium for distributing information materials for increasing disaster awareness.

Gun turret automatic weapon control system design and realization

Basically weapon is something that can protect themselves and groups of threats or interference. The development of weapons of today are increasingly sophisticated with the advancement of technology and information. Therefore, researchers are racing to develop technologies gun turret for defense and security interests that can be placed in tanks, armored vehicles, warships, and military aircraft. In this paper propose a gun turret that is able to follow the direction of motion of the target wherever move. So after the detection of the gun turret will move wherever the target will move either to the right, left, up, or down. The use of two DC motors mounted on each piece of coordinates, ie first at the coordinates pitch and second at the coordinates yaw as well as to measure the angle with the installation of two potentiometers, each one at the coordinates pitch and one at the coordinates yaw then tracking the target use one camera located above the gun turret. A PID controller is used to control the gun turret when tracking objects. By experiment, the proposed approach then verified.

DNA: Web-based content posting interaction in pervasive display

In the case of natural disaster, there is a need for an integrated information presentation system. The increasing number of public display could be used as a medium for that system. But currently, public display is developed as a closed system. Where only display owner could post content on it. It needs to move to a more open system that allows people contribution beyond display owner. In this paper, we proposed an interaction flow that could foster posting content activity in a system called DNA (Digital Notice Area). We tried to be similar to posting content activity in the traditional “Public Notice Areas”. From our preliminary research, we found that our proposed interaction has a SUS score of 69.25. Which means that it is acceptable but still need improvement.

Optimal cone of relative position acquisition module of multi mobile robot

The research of position acquisition of robots has been done in many years. The robots need to know the position of each other in order to be able to coordinate for task to be done together. The conventional method of position sensing is using odometry by utilizing encoder attached to the wheel which depends on non-slippery condition. This method results unexpected errors in most cases. Trilateration is one of the solution of the problem and the simple method of calculating position of multi mobile robot. This paper proposes an optimal cone of relative position acquisition module of multi mobile robot based on trilateration method and, by experiment, the proposed module was verified. The mean of accumulation error is 6.772 cm, which is practically appropriate for most swarm mobile robot navigation and coordination.