CBBR:enabling distributed shared memory-based coordination among mobile robots  

作　　者：Xue JIANG Yu HUANG 

机构地区：[1]State Key Laboratory for Novel Software Technology,Nanjing University,Nanjing 210023,China

出　　处：《Science China(Information Sciences)》2016年第8期19-35,共17页中国科学（信息科学）（英文版）

基　　金：supported by National Basic Research Program of China(973)(Grant No.2015CB352202);National Natural Science Foundation of China(Grant Nos.61272047,91318301,61321491);partially supported by Tencent,Inc.

摘　　要：Coordinating mobile robots are widely used in commercial and industrial settings to fulfill various tasks. However, to program the coordination among mobile robots is challenging. A coordination framework is needed to shield the programmer from handling low-level details of robot control and communication, while supporting flexible and cost-effective coordination at the same time. The coordination framework should also be able to well coexist with the underlying robot control. To this end, we propose the Coordination-enabled Behavior-Based Robotics （CBBR） framework. CBBR employs Distributed Shared Memory （DSM） to support coordination. The shared memory illusion built by the DSM greatly simplifies the coordination logic. Moreover, the flexible access patterns of the DSM and the rich consistency semantics of the DSM reads and writes enable flexible and cost-effective coordination. With the coordination support from the DSM, CBBR naturally extends the classical Behavior-Based Robotics （BBR） for robot control. From the perspective of robot control using BBR, the shared variables in the DSM act as the logical sensors capturing the status of coordination. The coordination algorithms are encapsulated into coordination behaviors. Thus, the physical environment status and the coordination status may trigger the physical and the coordination behaviors. The scheduling of both types of behaviors integrates coordination into robot control. We conduct a case study to demonstrate the use of CBBR. The performance measurements show the cost-effectiveness of coordinating mobile robots based on CBBR, in terms of time, space, and energy consumption.Coordinating mobile robots are widely used in commercial and industrial settings to fulfill various tasks. However, to program the coordination among mobile robots is challenging. A coordination framework is needed to shield the programmer from handling low-level details of robot control and communication, while supporting flexible and cost-effective coordination at the same time. The coordination framework should also be able to well coexist with the underlying robot control. To this end, we propose the Coordination-enabled Behavior-Based Robotics （CBBR） framework. CBBR employs Distributed Shared Memory （DSM） to support coordination. The shared memory illusion built by the DSM greatly simplifies the coordination logic. Moreover, the flexible access patterns of the DSM and the rich consistency semantics of the DSM reads and writes enable flexible and cost-effective coordination. With the coordination support from the DSM, CBBR naturally extends the classical Behavior-Based Robotics （BBR） for robot control. From the perspective of robot control using BBR, the shared variables in the DSM act as the logical sensors capturing the status of coordination. The coordination algorithms are encapsulated into coordination behaviors. Thus, the physical environment status and the coordination status may trigger the physical and the coordination behaviors. The scheduling of both types of behaviors integrates coordination into robot control. We conduct a case study to demonstrate the use of CBBR. The performance measurements show the cost-effectiveness of coordinating mobile robots based on CBBR, in terms of time, space, and energy consumption.

关 键 词：distributed shared memory behavior-based robotics COORDINATION behavior hierarchy behaviorscheduling 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]