Trutegra recently examined a large number of sensors and data points generated in automated crane systems, which are usually collected by a PLC for use in decision making, e.g., selecting automated moves. An overwhelming amount of data is accumulated and at times the data is changing rapidly, so much so that it may be missed by an operator simply looking at an HMI screen or a technician looking at the PLC code. Trutegra has developed a very sophisticated and extensive data collection system and is presently making it standard for all their projects, providing a “show and tell” at the end of the project with the option to purchase the system. The data collection system is much like a data historian, capable of monitoring a huge number of tags at a very high speed and cataloging the data in the server to be made available in a report.
This capability can result in the observation of sensor transitions during the commissioning of a system at a much faster rate than is obtained with an HMI screen. Any intermittent tripping or I/O triggers that occur can be captured, which is very helpful to the troubleshooting effort. Looking at the data as a visual representation, either a graph or a frequency of occurrence, adds another layer of interpretation and allows one to see “the bigger picture”. Trutegra has observed that with their new system it is possible to get a better understanding of how equipment is being used. As an example, they have shown that claims of 80-90% utilization of a crane, and therefore the need for a second crane to keep up with production, are untrue when all the data is correctly analyzed.
The manner in which cranes are being used by the operator is also revealed by this new system. Questions arise that can be answered:
- Are they traveling at full speed?
- Do they search back and forth to hunt for a position?
- Do they bump equipment into the walls?
- Do they frequently stop abruptly?
In some instances, it has been found that the automation capability was turned off and the crane operated manually, presumably since the operator did not trust automation.
The data collection package can give insight into things that are not usually visible and therefore not detected. Off-machine analysis, such as machine learning algorithms, can indicate that a slow trend and rise of wheel bearing temperatures, or the higher torque on the gearbox over a specific period of time points to an impending failure. There are other examples of the monitoring of varying temperatures, motor temperatures, and oil life, to estimate wheel life expectancy. These predictions are now based upon how the crane is being used in contrast to just considering, “how long has it been operated?” These analyses offer the ability to make a realistic estimate of the performance of the crane and of the components therein.
The data collection system provides information that can eliminate the needless replacement of parts, as well as avoiding a failure that results in downtime. The replacement of parts, when a pending failure is identified, can then be carried out as a planned outage or planned maintenance and not impact negatively upon production.