The large telescopes have a diameter of 9.4 meters that need a coating process to cover the mirrors surfaces with a thin layer of highly reflecting aluminum material. This coating ensures that nearly 100% of the photons from celestial objects impinging on the mirrors can be recorded by the sensitive telescope instruments. Due to the unavoidable deterioration of this reflective layer by progressive oxidation and abrasion by airborne particles, the mirrors must be recoated once a year.
The coating process is controlled by a Step5 Series Siemens PLC with WinCC as the operator interface. It was decided to add a Historian into the system, since data trending is an important aspect in any control system. In special situations when things go wrong or are otherwise not behaving properly, the data trending system remains the best diagnostic tool for figuring out why. One requirement of this project was to have minimum interference with the actual configuration and reduce the impact in the PLC load. Because of this, the decision was made to have the OPC Server already available in the WinCC application act as a “middle man” between the WinCC tag list and the Historian.
The first step was evaluating different OPC Historian solutions. The Canary Historian from Canary Labs was chosen because of its price/performance and other features that were discovered during the evaluation period. Some of these features were the ability to include comments in the trends during coating, and the ability to manage the Database in a batch mode, which was perfect for this process application. Also, additional data could be integrated into InfoLink using the scripting capability of the Canary products.
The Canary Historian and InfoLink were installed on a virtual machine under Mac Mini with OSX 10.5. The Canary tools monitor current and historical data, letting users cover all the different operation scenarios. The main advantage to having a real-time Historian is the volumes of data it can handle while allowing the user to analyze it easily.
By looking at the real-time trend plots, like temperatures, pressure, position, etc., the operator can follow the behavior of the process while identifying immediate spikes and transient data to find the cause of a fault and make a more informed decision. The operator can also notice when variables drift out of the operational limits by viewing trend plots. This allows the operators to be more familiar with the ideal trend characteristics of the system behavior and comprehensive monitoring of the Telescope and Instruments. Using the play-back feature available in the Canary tools, the operators can also reproduce a complete batch coating.
During troubleshooting, the engineers can visualize real-time trends to get a better understanding of automation and control loops, allowing for fine-tuning of the controllers. They also can model the process and optimize its control using additional tools available from Canary. The historical data allows engineers to define and monitor the Key Performance Indicators for each Telescope and Instrument, carrying out complex data processing while performing preventive maintenance based on actual usage or on statistical indicators allowing a proactive strategy.
The Figure shows part of the graphics interface developed during this project. This interface was built using the product InfoLink from Canary. The trend is in logarithmic scale and displays the chamber pressure during a complete coating process. The left side indicates the energy consumed during the process.
For more information about OPC solutions from Canary Labs, visit www.canarylabs.com.
