Open Software Structure for Controlling Industrial Robot

FLAVIO ROBERTI; SORIA, C.; SLAWIÑSKY, E.; MUT, V.; CARELLI, R.

Robot Manipulators, Trends and Development

In-Tech

Lugar: Viena; Año: 2010; p. 497 - 520

Automatic control has become an important part of the modern industrial processes. Progress both in basic research as applied to automatic control, provide a way to obtain the optimum performance of the dynamical systems, improve the quality and reduce the costs. Robotics, as a part of automatics, represents nowadays an important research area, and it has an essential role in the productive modernization (UNECE and IFR, 2005). The inclusion of industrial manipulators in the manufacturing process allows obtaining better and cheaper products. Therefore, the development of an open software structure for the industrial robots controlling is a very important objective to be achieved (William, 1994), (Frederick and Albus, 1997). The main characteristic of an open software structure for robotics applications is the interface that relates the components of the robot with the basic internal structure. In industrial area, one of the most important works was developed in the framework of the European project OSACA (Open System Architecture for Control within Automation Systems). Similarly, significant contributions were reached in Japan through OSEC (Open System Environment for Controllers) under IROFA (International Robotics and Factory Automation Center), (Sawada and Akira, 1997), and in the United States of America through OMAC (Open Modular Architecture Control). The objective of all these research projects is to develop an open control system including the reference model of the components, the general application interface and the structure so that all the components work together. So far manufacturers do not work together to develop standard control software that could be applied to any industrial robot. On the other hand, several commercial software packages, that run under Windows, for mobile robots can be found. Among the best known ones, Advanced Robotics Interface for Applications (ARIA) is used in the robots manufactured by Mobile Robots Inc., BotController software were developed by MobotSoft and it is used for the well known Khepera and Koala robots. Even when these software packages are powerful and have many benefits, they can be applied only to the robots that were developed. The main objectives of this chapter are the development and the implementation of an open software structure with reusable components, which works as a link between the hardware of an industrial robot manipulator and its control algorithm in order to implement these control algorithms with minimum efforts. Having this kind of software structure is very useful for researching and teaching in robotics as well as for industrial applications. The software structure runs under QNX Real Time operating system (Krten, 1999), and can be used for a large number of industrial robots. With the aim of achieving the raised objectives, the developed system is compound by two different programs. First one is the responsible for the sensors’ data acquisition and sending the control action to the servos. This program uses a shared memory block to save the data obtained from the sensors and to get the control action to be sent to the servos. In the second one runs the control algorithm. This program, similar to the first one, uses the same shared memory block to get the sensors’ data and to save the control action to be sent to the servos. This way, the control algorithm execution is isolated from the signals transmission between the software and the robot’s hardware, allowing a time and efforts reduction in the implementation of different control algorithms. Then, two different controllers have been implemented in order to evaluate the performance of the proposed open software structure, applied to the SCARA robot manipulator Bosch SR-800. First, a classical PD (proportional-derivative) controller is used to allow the robot to achieve a desired position on the workspace. This controller uses the position information from the encoders of the robot. Finally, an advanced passivity based visual servoing with “eye-in-hand” camera configuration (Weiss et al., 1987) is implemented to allow the robot to reach a position relative to some static target. Additionally, finite L2-gain for the passivity based control system is proven when a moving object is considered, allowing the robot to track the moving target with L2-gain performance. Experimental results for both, the classical PD controller and the passivity based visual controller are presented in order to show the good performance of the proposed open software structure when it is applied to industrial robot manipulators. This chapter is organized as follows. Section 2 describes the used industrial robot manipulator. Section 3 presents the open control software developed. Section 4 comments the control strategies used to evaluate the software structure and shows the experimental results. Finally, Section 5 presents same conclusions of the work.