How to use the Reliability Dashboard

Data types:

• Direct measure – a direct measurement is a value that can be trended that is a direct reflection of the machine health/reliability for example:
  • Vibration
  • Temperature
  • Pressure
  • Flow
  • Torque

• Indirect measurement – indirect measurement is a calculated value that can be trended and can be used to create a work order from, for example:
  • Filling valve – calculation to look at the waste or rejects being generated compared to the other filling heads under certain process parameters.
  • Capper head - calculation to look at the waste or rejects being generated compared to the other capping heads under certain process parameters.

Processing:

• Simple Parameter Processing: When monitoring data parameters for reliability it is normal to apply an upper and lower alarm limit on the trend. The algorithm is then a simple setting of the limit to trigger action. The alarms can be:
  • Rate of change (how fast is something changing)
  • Upper limits (normally and warning and alarm level)
  • Lower limits (normally warning and alarm level)

• Parameter processing with process effects: A single parameter for reliability can usually be affected by a number of process parameters. Therefore, the ML algorithm needs to understand how other process parameters effects the reliability parameters before setting the alarm bands. This should mean however that the alarm limit can be much tighter to what is expected.