The Sight Machine platform focuses primarily on data ingestion related to transactions between assets and materials during the production process. This data can be broken into a number of different data types. The following table provides examples of how the data types enable the use cases.
Data from sensors, programmable logic controllers (PLCs), scanners, machines, etc., created during the production process. For example, this could include pH readings from a sensor for a fermentation vat or scanner moisture percentage readings for a reel of paper. The first step in making process data valuable for analysis is turning the timeline of data on an asset or assets into discrete events. Process data can further be broken down into:
Productive: Data associated with the production process while material is moving in-process. This is the time in which the machine is doing its value add. This data can be in the form of a categorical indicator, binary, or even inferred from other sensor readings. For example, a packaging machine is in-process when the filling sensor readings are not null.
Non-Productive: Process data during non-productive machine use (either planned or unplanned). This is the time in which the machine is down. Again, this data can be in the form of a categorical indicator, binary, or even inferred from other sensor readings. For example, a welding machine is non-productive when the plasma sensor readings are 0. Often these readings will be reduced to a reason code to be able to group like downtime events for analysis.
Typically, data from test results and measurements associated with the material during or after a process. For example, this could include lab inspection test data for color and shine after the production of a reel of paper or an automated seat-back recliner test after the assembly of a seat frame.
Data that tracks the material flow through one or more asset. For example, serial numbers on individual packages or a lot identifier per batch of drug produced can be associated to track a vat’s contents though a pharmaceutical tablet press and packaging process.