From the point of view of information theory applied to the robotics system, network and interfaces oriented to the network in fact may act as a method and medium for dynamic resource allocation, i.e. the optimization of the large number of components in a complicated robotic system, and specially, in a workcell for complex and flexible tasks.
It expands the function and efficiency of a manufacturing unit as well as reducing the cost and time for function change, reconfiguring and tools preparation. It is a extension of emerging trends of agile manufacturing, parallel engineering and flexible manufacturing in the mechanical and robotics domain, which integrate the computer, information theory and technology and forecast future development.
There are two levels of network and interface in the robotic system. One is the network and interface at the manipulator level, the other is the network and interface at the workcell level. Both provide essential methods to implement modularity, reconfigurability and fault tolerance in the robotic system. The former is also called internal network and interface.
It provides communication ports, routes and protocols for signal exchange among different components inside the individual manipulator, such as actuators, controllers, internal sensors and processors. It implements modularity, reconfigurability and fault tolerance at manipulator level, that is, the manipulator structure and functions can be changed by replacing the ill-functioned components, add extra components and remove some components without affection to the performance of the system.
The counterpart in the workcell implements the similar functions, but with a broader scale and at a high level. The components in the workcell include individual manipulators, intelligent control, communication and management nodes, external sensors and PLCs.
It is the function of the network to make a physical and logical connection among the different components and resources in the workcell. It provides management, coordination, physical route for the exchange of Robot Control Command, External Sensing data, Error Report, Network Management data, Expertise System data and other command or data, to implement task simulation, task cooperation, workcell configurating and reconfigurating. It responds in real time with a very reliable transmission.
The generalized interface provides a standard port which masks all the variations of the port among the different workcell components. This kind of standard interface is expected to have the property of intelligence, that is, it can identify itself to the network on one side, adapt to the corresponding communication protocol (built in protocols) with components on the other side, implement plug-and-play and self error monitoring and reporting. Furthermore, interface responds to the data error-check, data integration and data translation. Our concept of communication of interface and communication structure is on the way to develop.
|