Using the object paradigm for distributed real-time systems

One of the primary features of almost all physically distributed software is that it has to deal with the unpredictable nature and complexity of the physical world. For example, it has to detect failures and undertake recovery procedures, it may have to reconfigure itself dynamically, and so on. In effect, all such distributed systems are de facto real time systems. As more and more software systems become distributed, the issue of architecture becomes more critical. A good architecture minimizes complexity and increases the likelihood of run time controllability. It is also key to its ability to undergo evolutionary change. Consequently, the ability to model complex layered architectures in a formal and clear way is also important. The structural modeling concepts originally defined in the ROOM (real time object oriented modeling) language provide an excellent base for this purpose.