Automated Sources for Waste

As with downtime, planning for automatic waste detection starts with auditing the available signals. There are several types of signals we are looking to automatically drive waste calculations.

  • Directly measured waste streams with or without reject reasons

  • Directly measured raw material inputs

  • Inferred raw material inputs (using combined signals and/or conversion factors)

  • Directly measured production outputs

  • Inferred production outputs

  • Signals implying product or configuration

Frequently, knowing the product or configuration of the line is important to correctly interpreting raw material and production signals. For example, the product code may help to infer how much finished product is expected from a given amount of raw materials.

How To Prepare The Control System For Automatic Waste Detection

  1. Locate all signals that measure production

  2. Determine what conditions make each signal valid or invalid. This could be based on configuration, product code, or by direct feedback of another signal.

  3. For material balance type models, determine the conversions between inputs and outputs expected for each product the line will run.

  4. For direct measurements, determine the behavior of the signals. How fast do the values change? When do they change?

  5. Where timing is critical, supplement logic in the control system to accumulate or filter values

  6. Collect all signals into the historian at the highest resolution possible.

What Can Be Automated? 

  • Waste Times

  • Waste Production Event

  • Cause Location, Or Location Of Material Balance

  • Waste Type And Measurement

  • Default Reasons Can Be Model Specific

Examples:

  • Photo eye counts two input streams

  • Photo eye counts output stream

  • Active product determines expected conversion of inputs to output

  • Model determines Waste Type, Measurement, and Amount

  • Model Defaults Reasons

Example Types of Downtime and Waste Signals