FROM MOBILE ROBOTS TO INDUSTRIAL AUTOMATION Le Vu Ha, Pham Duy Hung, Ha Khanh Linh, Tran Quang Vinh, Dang Anh Viet, Nguyen Thi Thanh Van Automation and Robotics Laboratory College of Applied Sciences and Technology Vietnam National University, Hanoi halv kcn@vnu.edu.vn ABSTRACT In this paper we present a model for research at a univer- sity laboratory whose primary function is twofold: con- ducting research on advanced technologies in automa- tion and robotics, while serving as a practical laboratory to provide our technical staff and students with work- ing experience on industrial automation systems. The model features a systematic process of developing con- trol systems hardware and software which comply with industrial standards, using home-made experimental plat- forms like mobile robots as research and development testbeds. It allows us to combine two tasks – performing advanced researches and developing industrial-standard systems and components for training purposes – in an ef- fective way, given our limited resource. 1. INTRODUCTION Our long-term goal is building this laboratory into a mod- ern research and training facility in automation, especially robotics and intelligent systems. As a newly-established laboratory, it may take years to develop its infrastructure to meet that goal. On the other hand, we can not afford to wait until the facility is fully equipped and staffed in order to start functioning. Despite current lack of resource, par- ticularly in personnel and space, we have been deploying research projects on various advanced topics, including distributed control, virtual control systems, and vision- guided robot navigation, while building up systems and components following common industrial standards for the industrial automation division of our laboratory. The process also helps our staff and students to obtain valu- able experience on developing industrial automation sys- tems and makes way for our research products to easily integrate into industrial systems for real-world applica- tions. We have been looking for a model of operation to carry out the tasks effectively within our limited means. In the next sections we will present a model we have come up with and show how research projects are con- ducted under its guidance through an example, a project on robot navigation techniques. A mobile robot is built to experiment navigation algorithms. The robot is also used to test implementations of control networks includ- ing network hardware and communication protocols. Be- cause the example itself can draw attention from read- ers who are interested in control networks, we will try to present it as detailed as possible within the scope of this paper. 2. INTEGRATED MODEL OF RESEARCH AND STANDARDIZED CONTROL SYSTEMS DEVELOPMENT As the title suggested, the main characteristic of the model is the integration of research processes with the develop- ment of industrial-standard control systems hardware and software. Why are we talking about control systems in- stead of automation systems? An automation system is a complex of several subsystems, including manipulators, sensors, and control subsystems. Building such systems requires various fields of expertise, from mechanical en- gineering to electrical engineering, networking, artificial intelligence, and software engineering. Since our back- ground mostly lies within the areas of electrical engineer- ing and computer science, it is more appropriate for us to focus on industrial control systems rather than on all as- pects of automation. On the other hand, this does not limit the scope of our research. The model is characterized by the following features: • Integration: as we mentioned above, this model combines research processes and the development of control systems components into a same project. Research topics may come from many active re- search areas of automation and robotics, such as robot vision, sensor fusion, mobility mechanisms, and etc. The control systems components are built into experimental platforms, used for evaluating re- search results. It allows us to concentrate our re- source for each project while still accomplishing both tasks at a same time. • Reusability: the experimental products (both hard- ware and software) should be designed in such a way that their modules can be easily extended and integrated into other systems, desirably industrial systems. Libraries of control hardware and soft- ware components are maintained and continually