A Common Cell Control Software
In this section, fine and common implemented cell control software are developed, that is based upon a transition table, to show the restrictions of the system which can be simply resolved via the utilization of object oriented control architecture. There are three categories of software modules as: a controlling manager, action tasks that perform exact actions in the system, and support functions that interface to the cell devices. Messages are passed through a set of mailboxes from the control manager to the action task to execute the suitable tasks as dictated via the state table. All action tasks, in turn, sends back a message while its task is finished. Each of these tasks has specific responsibilities.
Various restrictions exist in this cell control software. Even if functional, the original code is somewhat complicated to maintain since it is implemented in a procedural language. The inter actions of simultaneous processes are very hard to grasp based upon a reading of a procedural code. The other issue is the level of redundant programming in the software since same sections of codes are required in various action tasks. Throughout software development a continual trade-off exist among the level of hard coded, redundant modules and the degree of parameterization employed for all modules. By using various similar, but particularly written modules identifies the logic readability of the whole system, whereas use of an individual, heavily parameterized module decrease the code but raise execution time.
An additional and associated limitation is the complexity of adding latest devices to the cell that needs revised software modules or that latest module is coded and connected to the system. Possibly the most serious deficiency of the system is which it restrictions the cell to having an individual part on the bed of the router at some time. Machine tool technology is such as multiple part placements are possible to attain enhanced productivity. Multiple part processing is possible by using state table approach but explosive development in the size of the table as elements and or machines are added in a problem.
At last, more robust error diagnosis and resolution functionality are required. In the unique implementation, there are various error conditions from that manual recovery is needed. Conversely, with an enhanced architecture several of these manual recoveries could be automated.