Tom Nolan, ProSys
For years, a great deal of automation was custom made with line by line source code. The automation has proven itself in terms of value, but the monolithic architecture can be difficult to maintain, leverage to other applications or migrate from obsolete systems. For some obsolete legacy systems, such as Dow Chemical’s MODV, there are few practitioners left who can effectively reverse engineer the existing code. These practitioners are also, as a rule, not well versed in newer platforms and a modular architecture. The remaining practitioners are very near retirement age and younger control engineers may not see mastering the intricacies of an obsolete platform as a great enhancement to their resume.
The modular format has some proven benefits and can be utilized into a software product line type approach. Software product lines have demonstrated order of magnitude improvements in:
- Time to market
- Cost reduction
- Enabling mass production and mass customization of software products
Effective use of a modular approach has demonstrated projects completed in one-third the time as compared to conventional methodology, with one quarter of the resources and a ninety-six percent reduction in bugs in the final software product. This results in lower cost:
- Brown field projects
- Green field projects
Beyond cost, another major benefit is risk reduction in brown field, green field and migration projects. Project risks include:
- Delays in start-up
- Problems with product quality
- Safety and environmental incidents
- Accounting or environmental reporting problems
- Cost escalation
- Ability to get and keep resources with the right skill set
- Ability to migrate from an obsolete system before failures or lack of support is available
Using a modular approach and reverse engineering “proven in use” functionality greatly reduces risk.
The functional specification enables consensus on what will be built before coding and simulation are done. As a rule, a problem found earlier in the work process is ten time less expensive to fix than one that is found later. Then, the functional specification is the basis for testing, training and commissioning with accountability all the way back to the original source code.
The core assets also improve the return on investment since they are available to be leverage to any other type of project.
Since the specification and later the software are assembled from “proven in use” applications that have mined the functionality from the legacy code, the risk in quality and safety performance or in delays in start-up are greatly reduced.
Risk is also greatly reduced in terms of resources. Fewer resources are needed. Even fewer highly skilled resources are needed for reverse engineering and core asset development. Instantiation and development can be achieved with less skilled / experienced resources. In the event of turn over, which is likely with millennials moving around and boomers retiring, this layer of resource can be easily trained with a good set of standards.
Time risk is reduced since the time to complete a high quality project can be a third of what it might normally be, the chances of flying an obsolete system into the ground are greatly reduced, as well that the chance of missing any start-up milestones that may be linked in with multiple plant turnarounds.
There is software available, whose work process and reverse engineering capability bridges the gap by reverse engineering the functionality of legacy code and refactoring it into a modular architecture. This is done in a format that accounts for all legacy code, so it’s understood by operations folks, legacy platform and targeted platform practitioners. It greatly reduces the required number of practitioners with knowledge of the legacy system for reverse engineering while improving the quality of the final product. One such software is The Legacy Refactor, available through ProSys.
Bottom line, a modular architecture can greatly reduce cost and risk while improving quality and consistency.