실내 환경에서 운영 가능한 RFID 기반 멀티 로봇 관리 시스템

안상선(Sang-Sun An),신성욱(Sung-Oog Shin),이정욱(Jeong-Oog Lee),백두권(Doo-Kwon Baik) 한국컴퓨터정보학회 2008 韓國컴퓨터情報學會論文誌 Vol.13 No.6

오늘날 멀티 로봇에 대한 연구는 단순한 싱글 로봇들의 효율적인 운영을 넘어 공간탐색의 효율성 극대화 및 넓은 공간에서 각 로봇간의 중복작업 및 충돌 회피를 위한 부분에 집중되어 있다. 이러한 멀티 로봇을 효율적으로 운영하기 위해서는 각 싱글 로봇을 제어하고 효율적으로 작업을 할당 할 수 있는 관리체계가 필요하다. 이에 본 논문에서는 RFID를 기반으로 각 싱글로봇의 탐색공간을 효율적으로 할당함으로서 싱글로봇간의 중복 탐색을 최소화 할 수 있는 멀티로봇관리시스템을 제안한다. 또한, 탐색 작업의 완료 보장과 탐색 성능의 향상을 위하여 장애가 발생한 싱글 로봇을 탐지하고 대체 할 수 있는 고장 허용 기법을 제안한다. 제안한 시스템에서는 정확한 위치 파악이 힘들었던 기존 중앙 서버의 단점을 RFID 시스템과 홈로봇을 활용하여 극복하였다. 지정된 홈로봇은 각각의 싱글 로봇을 효율적으로 관리하며, RFID 태그의 위치 정보를 활용하여 각 싱글로봇에게 최적의 탐색 공간을 할당 한다. 제안하는 멀티로봇 관리 시스템은 공간 할당, 위치 추정 기법, 맵 생성 기법(Localization및 Mapping)을 효율적으로 수행하기 위해 RFID를 기반으로 하며, 싱글 로봇 시스템과 비교하여 시스템 운영의 효율을 극대화할 수 있을 뿐만 아니라 각 싱글 로봇의 상태와 주변 상태를 고려한 고장 허용(fault tolerance)을 제공함으로써 로봇 운영의 신뢰성을 보장할 수 있다. 본 논문에서는 시뮬레이션을 통해 제안한 시스템을 적용한 멀티로봇 시스템과 기존의 멀티로봇 시스템의 탐색 소요시간 및 중복 탐색율을 비교하여 제안한 시스템의 효율성을 입증하였다. The multi robot operation technique has emerged as one of the most important research subjects that focus on minimizing redundancy in space exploration and maximizing the efficiency of operation. For an efficient operation of the multi robot systems, the movement of each Single robot in the multi robot systems should be properly observed and controlled. This paper suggests Multi Robot Management System to minimize redundancy in space exploration by assigning exploration space to each robot efficiently to take advantage of the RFID. Also, this paper has suggested fault tolerance technique that detects disable Single robot and substitute it by activated Single robot in order to ensure overall exploration and improve efficiency of exploration. Proposed system overcomes previous fault that it is difficult for central server to detect exact position of robot by using RFID system and Home Robot. Designated Home robot manages each Single robot efficiently and assigns the best suited space to Single robot by using RFID Tag Information. Proposed multi robot management system uses RFID for space assignment, Localization and Mapping efficiently and not only maximizes the efficiency of operation, but also ensures reliability by supporting fault-tolerance, compared with Single robot system. Also, through simulation, this paper proves efficiency of spending time and redundancy rates between multi robot management applied by proposed system and not applied by proposed system.

Multi Agent Robot System(MARS)의 Robot Agent 간 정보교환을 위한 네트워크 프로그램 구현

고광은(Kwang-Eun Ko),이정수(Jeong-Soo Lee),장인훈(In-Hun Jang),심귀보(Kwee-Bo Sim) 한국지능시스템학회 2007 한국지능시스템학회논문지 Vol.17 No.5

홈서버와 가정용서비스 로봇, 다양한 Device를 포함하는 홈네트워크 시스템을 이용하여 가정환경 내에서 발생 가능한 다양하게 분산된 임무들을 수행하는데 있어서 Multi-Agent 시스템을 응용하는 것은 효율적이라고 알려져 있다. 이러한 시스템에서 휴먼 인터페이스와 물리적 서비스 제공의 중추 역할을 하는 것은 바로 지능형 서비스 로봇이다. 따라서, 기존의 Multi Agent System을 응용하여 Multi Agent Robot System을 정의할 수 있다. 가정이라는 ‘open’된 특성의 환경에서 Multi Agent Robot System이 사용자에게 좀 더 효율적인 서비스를 제공하기 위해서는 존재하는 모든 Agent 간의 유기적인 정보교환과 협력이 필요하다. 이것을 위하여 자율구동하며 사용자에게 물리적 서비스를 제공할 수 있는 다수의 로봇들을 Agent로 정의하고, 로봇 Agent간의 통신과, 로봇 이외의 Agent로써의 역할을 감당할 수 있는 홈서버 등의 Agent와 로봇 Agent들의 통신을 사용자 interface에 직접 표현할 수 있는 시뮬레이터를 설계, 제시한다. Using home network system including home server, home service robot, a variety of device, it is generally known that application of Multi Agent System for performing variously distributed process that can be occur in home environment, is efficient method. In this system, it is intelligent service robot that a key of human interface and physical service offer. Therefore, using application of established multi agent system, we can defined Multi Agent Robot System. In 'open' home environment, between all agent data interaction and cooperation are needed for Multi Agent System offer to user that more efficient service. For this, we focus our attention on define as agent that can autonomic drive and offer to user that physical service robots and, design, suggest the simulator can display that between robot agents communication or between other agents, like home server, and robot agents communication information to user interface.

A Multi-Robot Task Planning System Minimizing the Total Execution Time for Hospital Service

Yong-Hwi Kim,Byung Kook Kim 제어로봇시스템학회 2010 제어로봇시스템학회 국제학술대회 논문집 Vol.2010 No.10

A multi-robot task planning system is developed to minimize the total execution time for autonomous service robots in hospitals, which is essential for ‘autonomous generation of task commands’ as an approach toward human-friendly human-robot interaction. In traditional approaches, a task planning system based on STRIPS (Stanford Research Institute Problem Solver) representation is widely used. Also the temporal planning is one of the issues in task planning area, since it still suffers some problems for single robots, and also it is not easy to apply to multi-robot environment yet. For our objective, we propose a new representation, named variable representation, based on STRIPS representation for specifying states and actions. Then we apply the A* algorithm for iterative search and extend for multi-robot task planning system. In order to guarantee safety, the multi-robot task planning system needs to check robots’ collisions and should avoid robots’ collisions. We propose collision check and avoidance algorithm, and integrate to the multi-robot task planning system.

비무장지대 다족로봇의 적응형 보행패턴 도출을 위한 강화학습 연구

김재환(Jaehwan Kim),김종환(Jong-Hwan Kim) 국방로봇학회 2024 국방로봇학회 논문집 Vol.3 No.1

This paper presents the challenges posed by the Demilitarized zone in the Korean Peninsula, which is characterized by rugged mountainous terrain and the presence of hidden mines, making it a hazardous environment for traditional wheeled mobile equipment and robots. In such areas, the use of multi-legged robots with lower ground contact is more suitable. However, there is a lack of research on robot control mechanisms capable of responding quickly to unexpected situations like leg loss in multi-legged robots. This study focuses on developing control models and reinforcement learning strategies to enable rapid adaptation to leg loss during the movement of multi-legged robots designed for mine detection. The MA-POCA(Multi Agent-Posthumous Credit Assignment) algorithm, one of multi-agent algorithms in reinforcement learning, is employed for this purpose. Additionally, a simulation model and environment are created using Unity and the ML-Agents library to facilitate the rapid learning of unforeseen damage scenarios during robot operations in the DMZ. To address this challenge, each agent within the multi-agent group is assigned to control one leg of the multi-legged robot and is trained as an independent agent using CTDE (Centralized Training via Decentralized Estimation). This approach overcomes the limitations of traditional supervised learning methods and is empirically shown to be effective. Notably, the article highlights that the gait developed using the MA-POCA algorithm, focusing on control and reinforcement learning for lost legs during walking, distinguishes itself from other algorithms primarily focused on multi-legged robot locomotion. This unique approach allows for adaptation to critical situations that may occur during mine-clearance operations in the DMZ, suggesting potential improvements in mission efficiency, particularly in rapidly changing situations.

Evaluation of Robot Selection Algorithms for Organizing a Robot Cluster in Multi-robot Systems

홍영식,정진우,H. W. Koo,J. H. Jung 동국대학교 정보융합기술원 2008 International Journal of Assistive Robotics and Sy Vol.9 No.3

Grouping robots is one of the most important issues in multi robot systems. Most studies on grouping have the objective of realizing cooperation and coordination between robots for solving specific problems. This set of robots is called a robot cluster. This paper evaluates robot selection schemes employed for a robot cluster to perform a specific task. We evaluate simple robot selection methods through an analysis of the time used to organize a cluster by Network Simulation 2 (NS-2). The obtained results along with fundamental data are used to develop an efficient robot selection algorithm. In addition, problems related to organizing the cluster are discussed

Behavior tree driven multi-mobile robots via data distribution service (DDS)

Seungwoo Jeong,Taekwon Ga,Inhwan Jeong,Jongeun Choi 제어로봇시스템학회 2021 제어로봇시스템학회 국제학술대회 논문집 Vol.2021 No.10

We propose multi-task programming for multi-robots via a behavior tree (BT) as a mean to reduce programming complexity of multi-mobile robots. For multi-mobile concurrent control, robot sensors or actuators data are shared via data distribution service (DDS). In addition to the advantages of DDS, the BT is deployed to control and monitor multi-mobile robots, The BTs are inherently modularized and reactive to the environments, which enables a flexible utilization of multi-mobile robots. With the help of Robot Operating System (ROS2), The BTs are implemented with ROS2 nodes on the DDS that can be executed in both servers and clients for the task allocation. A simple experiment with 50kg, 100kg payload mobile robots built by Hyundai Robotics shows reliable performance with distributed BTs via DDS between servers and robots.

시설원예 작물수확을 위한 이종 복수 로봇의 작업효율 분석

최태용,박종우,김성중,신영식,김두형 한국통신학회 2022 韓國通信學會論文誌 Vol.47 No.10

Recently, a lot of research on green house such as smart farms to which the 4th revolution technology is applied has been conducted. Among them, robot technologies such as harvesting robots and transporting robots to move crops by human workers are being researched. Existing research is mostly focused on the realization of functions of a single robot. In this study, it is shown that when a plurality of harvesting robots and transporting robots are properly mixed for harvesting crops in green house, work efficiency is higher than using only a single harvesting robot. To this end, the harvesting speed and transportingspeed of the harvesting robot and the transporting robot are modeled in consideration of the facility environment, and based on this, the working speed of the robot group consisting of the harvesting robot and the group consisting of the harvesting robot and the transfer robot is derived. In addition, by analyzing the derived working speed based on the experimental data of the actually developed harvesting robot and transporting robot, it is shown that the overall working speed is higher when the harvesting robot and the transporting robot are properly mixed and cooperative work is performed between the robots. 최근 4차혁명 기술을 적용한 스마트팜과 같은 시설원예에 대한 연구가 많이 이루어지고 있다. 그 중 수확로봇과 사람 작업자의 수확물 이송을 위한 이송로봇 등의 로봇 기술이 연구되고 있다. 기존 연구는 대부분 단일 로봇의 기능 구현에 치중되어 있다. 본 연구에서는 시설원예에서 작물 수확를 위해서 복수의 수확로봇 및 이송로봇을적절히 혼용했을 때, 단일 수확로봇만 사용하는 것보다 작업 효율이 높음을 보인다. 이를 위해서 수확로봇 및 이송로봇의 수확속도 및 이송속도 등을 시설 환경을 고려하여 모델링하고, 이를 기반으로 기존의 수확로봇으로 구성된 로봇 그룹과 수확로봇 및 이송로봇 혼용으로 구성된 그룹의 작업속도를 도출 및 분석한다. 또한 도출된 작업속도를 실제 개발한 수확로봇 및 이송로봇의 실험 데이터를 바탕으로 분석하여 수확만을 위한 수확로봇과 작물이송만을 위한 이송로봇으로 구성했을 경우보다 수확로봇과 이송로봇을 적절히 혼용하고 로봇끼리 협동작업을 했을 때 전체적인 작업 속도가 더 높음을 보인다.

Experimental Analyses of an Efficient Aggregated Robot Processing with Cache-Control for Multi-Robot System

Abdul Jalil,Jun Kobayashi 제어로봇시스템학회 2020 제어로봇시스템학회 국제학술대회 논문집 Vol.2020 No.10

This study aims to develop a network architecture for low-cost multi-robot systems, in which robots operate with limited hardware capability and on-board computational power. Low-speed data processing and high latency time in communication are unavoidable issues in such a multi-robot system. To address those problems, this study proposes Aggregated Robot Processing – Robot Operating System (ARP-ROS) as a network architecture for low-cost multi-robot systems. This network architecture arranges a computing environment dedicated to data processing in a multi-robot system. In the proposed network architecture, most of sensor data obtained by each robot are transmitted to the computing environment via the communication network and processed there, and then control commands are sent back to the robots. For the data exchange in the network architecture, ROS2 node communication is employed because it uses Data Distribution Service (DDS) suitable for real-time distributed embedded systems. In addition, Cache-Control Algorithm (CCA) is introduced to the communication network for its performance improvement. This paper presents experimental analyses of the ARP-ROS with CCA in terms of latency time and data process failures.

군집 로봇 운용을 위한 상위 수준 프로그래밍 언어 조사

강우석,정은진,윤태준,허석행,하순회 한국정보과학회 2022 정보과학회논문지 Vol.49 No.11

Robots used for unmanned tasks in various fields have been developing in the direction of operating multiple robots rather than using only a single robot. Since swarm robotics aims to outperform a single robot through cooperative control of multiple robots, it can efficiently perform tasks such as building local maps and searching for survivors. Although developing and operating such swarm robot systems requires a skilled developer, high-level programming languages and frameworks that can support non-experts to easily specify and operate swarm robots are being developed. In this paper, we selected and introduced 11 swarm robot specification languages tested by simulations or using real robots. We compared the difference between selected languages from the viewpoint of software and swarm robot operation. Finally, limitations of swarm robotics languages developed so far and future tasks are discussed. 다양한 분야에서 작업을 무인화하는 데에 사용하는 로봇은 단일 개체로만 사용하기 보다는 여러 개의 로봇을 운용하는 방향으로 발전하고 있다. 군집 로봇 (Swarm Robotics)은 다수의 로봇을 협동 제어를 통해 단일 로봇 이상의 성능을 내는 것을 목표로 로봇 간의 협력을 통해 생존자 수색이나 지역 지도 작성과 같은 작업을 효율적으로 수행할 수 있다. 이러한 군집 로봇 시스템을 개발하고 운용하는 것은 숙련된 개발자를 필요로 하기 때문에 비전문가가 군집 로봇 운용을 쉽게 명세하고 운용할 수 있도록 지원하는 프로그래밍 언어 및 프레임워크가 개발되고 있다. 본 논문에서는 지금까지 연구가 이루어진 여러 상위 수준 군집 로봇 운용 명세 언어 중 실제 로봇이나 시뮬레이션을 통해 실행을 검증한 11개 언어를 선정하여, 이에 대해 간략하게 소개한다. 또한, 소프트웨어 및 군집 로봇 운용 관점에서 대상 언어들의 차이점을 비교하고, 현재까지 개발된 군집 운용 언어의 한계와 향후 과제를 논의한다.

Evaluation of Robot Selection Algorithms for Organizing a Robot Cluster in Multi-robot Systems

Y.S. Hong,J.W. Jung,H.W. Koo,J. H. Jung 한국과학기술원 인간친화 복지 로봇 시스템 연구센터 2008 International Journal of Assistive Robotics and Me Vol.9 No.3

Grouping robots is one of the most important issues in multi robot systems. Most studies on grouping have objective of realizing cooperation and coordination between robots for solving specific problems. This set of robots is called a robot cluster. This paper evaluates robot selection schemes employed for a robot cluster to perform a specific task. We evaluate simple robot selection methods through an analysis of the time used to organize a cluster by Network Simulation 2 (NS-2). The obtained results along with fundamental data are used to develop an efficient robot selection algorithm. In addition, problems related to organizing the cluster are discussed.