機器人應用系統開發：從設計、實作、測試、佈署、除錯、調適等程序涉及大量且複雜的方法、技術和工具，也耗費許多時間與人力成本。本研究提出一個新的平台，用以簡化並加速機器人應用系統的開發。我們基於MIAT嵌入式系統設計方法論，結合Web-based GPP編輯器(Grafcet Application ，GPP)、Grafcet虛擬機器(Grafcet Virtual Machine ，GVM)與機器人作業系統(Robot Operating System ，ROS)，建構出機器人應用系統開發平台。藉此平台，開發者能在遠端以Grafcet圖形語言快速描述GPP系統行為，再以編碼表形式將GPP傳送至遠端機器人控制器的Grafcet虛擬機器進行解譯，執行相對應的控制和感測程序。至於機器人控制系統架構，我們採用Modbus現場總線協定建立主從式通訊協定來實現分散式、多核心的機器人控制器，可滿足系統彈性化的擴充需求。在論文最後，我們以一個基於ROS的機器人應用系統開發實例來驗證本平台的可用性和性能。

A robot application system is typically developed through the processes of design, implementation, testing, deployment, debugging, and adjustment. These processes not only involve a variety of complex methods, technologies, and tools, but they also require considerable time and human resources. This paper proposes a new platform that simplifies and expedites processes for developing robot application systems. Based on a design methodology proposed by the Machine Intelligence and Automation Technology Laboratory for embedded system software, a Web-based GRAFCET Application (GPP) editor, a GRAFCET virtual machine (GVM), and a robot operating system (ROS) are incorporated to construct a robot application system development platform. System developers can employ the proposed platform to describe GPP system behaviors in a short time by using GRAFCET graphical languages at a remote end. The GPP system behaviors are then transmitted in the form of a code table to the GVM of remote robot controllers to decode the GPP and perform corresponding control and sensing processes. To develop a robot control system framework, the Modbus fieldbus protocol is applied to create a superior–subordinate communication protocol for achieving a distributed-control-based multicore robot controller that fulfills the flexible expansion requirements of systems. An empirical instance of robot application system development based on the ROS is cited in this paper to verify the usefulness and performance of the proposed platform.

[1] 走向多樣少量且系統化的生產 鞏固半導體產業全球競爭優勢, 2015; http://www.ey.gov.tw/news_Content2.aspx?n=F8BAEBE9491FC830&s=A09EFA1FB04C793A.
[2] 產銷模式（Production-marketing model）, http://wiki.mbalib.com/zh-tw/%E4%BA%A7%E9%94%80%E6%A8%A1%E5%BC%8F.
[3] O. Ogorodnikova, “Robot introduction in Human work environment. Developments, Challenges and Solutions,” Computational Cybernetics, 2007. ICCC 2007. IEEE International Conference on, 2007, pp. 167 - 172.
[4] Schmidt, et al., “Middleware R&D challenges for distributed real-time and embedded systems,” ACM SIGBED Review, vol. 1, 2004, pp. 6-12.
[5] ARM Ltd, http://www.arm.com/products/processors/index.php.
[6] M. Quigley, et al., “ROS: an open-source Robot Operating System,” ICRA workshop on open source software, vol. 3, 2009, pp. 5.
[7] A.M. Romero, ROS Concepts, 2014; http://wiki.ros.org/ROS/Concepts.
[8] DariushForouher, ROS rqt_graph, 2014; http://wiki.ros.org/rqt_graph.
[9] ROS robot simulators, http://wiki.ros.org/Robots.
[10] C.-H. Chen, et al., “The design and synthesis using hierarchical robotic discrete-event modeling,” Journal of Vibration and Control, 2012.
[11] C.-H. Chen, et al., “High efficient VLSI implementation of probabilistic neural network image Interpolator,” Journal of Vibration and Control, 2012.
[12] C.-H. Chen, et al., “A pipelined multiprocessor SOC design methodology for streaming signal processing,” Journal of Vibration and Control, 2012.
[13] V. Serifi, et al., “Functional and Information Modeling of Production Using IDEF Methods,” Strojniski Vestnik-Journalof Mechanical Engineering, vol. 55, 2009, pp. 131-140.
[14] 謝欣蓓, 「整合Grafcet虛擬機器的智慧型控制器開發平台」, 國立中央大學, 碩士論文, 2011.
[15] Modbus.org, MODBUS Protocol Specification, 2012; http://www.modbus.org/specs.php.
[16] G. Coulouris, et al., Distributed Systems: Concepts and Design (5th Edition), Addison Wesley, 2011.