CSense highlighted in MMS Mag - Wednesday, March 12, 2008

Rapid process troubleshooting means rapid process improvement

 

Today, in the world of process control and information, what is not lacking is information. In fact, its abundance can be overwhelming when specific information is required to address an emergency, to troubleshoot a process in difficulty or to improve the production bottom line. What's needed is a diagnostic tool that can help automation and control personnel "see sense" in their huge data repositories and that can point them in the right direction to eliminate the cause of problems and deficiencies.  By Jacques Ludik

 

About rapid process troubleshooting

Process variation such as deviation from targeted throughput and product quality is expensive.  Process deviation can also be hard to explain.  Most processes are complex and dynamic and to relate relevant process variables to process performance is not simple.  Furthermore, variation can be difficult to reduce, whether by understanding how to modify the operational philosophy, diagnostic solutions or advanced control. Reduction in process variance leads to far more cost -effective operations and drastic savings and the key to these improvements can be found in most process control historians. Yet, historian data, in general, is poorly utilised compared to the value that can be realised with model-based troubleshooting techniques.

 

"Rapid Process Troubleshooting" refers to the methodology and technology implemented to troubleshoot production problems rapidly and easily and to deploy solutions for throughput, quality, energy and data quality improvement. In today's competitive and cost-conscious markets, one no longer has the luxury of a choice whether or not to adopt rapid production problem diagnoses and cures. Ten years ago, a product's selling price was based on its manufacturing cost plus a profit margin. Today, competition pegs the selling price and the only option left for maintaining or improving the profit margin is to lower the product's manufacturing cost, and that means processes that run smoothly. 

 

The troubleshooting methodology allows users to identify the causes of a process problem and to develop an appropriate solution.  Data collected from the process is verified and validated using tailor-made data pre-processing blocks.  These contain dynamic modelling and data processing functions, which are used to build a blueprint of the solution. The models explain the causes of process exceptions as well as desired or unwanted behaviour. At each step of the solution blueprint, the software checks data quality and suitability before any processing is carried out. This ensures that estimations, decisions or control actions are made only when process information is reliable. 

 

By following the formal troubleshooting methodology that the CSense software provides, users can make sense of the complexity of their production or manufacturing processes and take appropriate action. Knowledge of what causes the process problem can be combined with operational knowledge in the deployment of real-time solutions.

 

Identifying the causes of KPI deviations

CSense Cause+ helps users define process deviation scenarios for KPIs and makes these Cause+ scenarios available to leading third party HMI/SCADA platforms. The underlying Cause+ model is fully automated and rules are automatically generated for the process specialist for each possible manipulated variable cause. Users simply need to fill in the associated actionable operator messages based on their practical knowledge of the situation. Furthermore, users are able to simulate the defined Cause+ scenario in the same CSense environment on the historical data, modify existing scenarios and simulate the modified scenarios again until they are happy to deploy the Cause+ scenario. 

 

Applications

Rapid process troubleshooting has been applied to process diagnosis, decision support, supervisory process control, materials accounting, optimisation as well as condition management of equipment and processes – in fact, anywhere it's necessary to determine the causes of variation in production KPIs, to increase throughput, yield or recovery or to improve quality and energy efficiency.  This technology can also be used to reduce or eliminate on-line mass and energy balances for improved materials accounting and production scheduling.

 

This technology in general and CSense in particular, has been proven on many troubleshooting and large-scale on-line solution projects in industries such as minerals processing, industrial minerals, pyrometallurgy, hydrometallurgy, steel, aluminium, ferrometals, paper and pulp, manufacturing, food and beverage, breweries, energy and chemicals.  Projects range from defect troubleshooting, process optimisation and stabilisation, process plant control, chemistry predictors, soft sensors and equipment condition health monitoring as well as data validation and production accounting. 

 

Conclusion

CSense Systems has developed a rigorous troubleshooting methodology embedded in its products that facilitates an accelerated Six Sigma approach. It is this methodology that guides users to rapidly discover new knowledge in historical data that explains the causes of production problems. Subsequently, this newly-discovered knowledge can be combined with existing knowledge such as rules, best practices and mathematical models, to deploy an Action Object (on-line process application) to solve and/or manage the identified problem.  In this respect, the software does not only add value to existing HMI infrastructure such as historians, SCADAs, and DCSs, but also to Management Execution System (MES) applications and, ultimately, to the bottom line. 

 

 

 

 

 

 

 

 

 

 

Author : Jacques Ludik - CSense

Date : Sunday, March 09, 2008