Edit Page

Configure the KPI service

📝The KPI service is currently provided provided through a separate Helm chart. If you want to use for your Rhize installation, get in touch.

This guide shows you how to configure the time-series you need for the KPI service. It does not suggest how to persist these values.

To learn how the KPI service works, read About KPI service. Example use cases include OEE and various performance metrics.

Prerequisites

Before you start, ensure you have the following:

  • The KPI service installed
  • An equipmentHierarchy is configured

Procedure

In short, to configure the KPI Service, the procedure works as follows:

  1. Persist machine state records to the EquipmentState table
  2. Persist quantity records to the QuantityLog table
  3. Persist job response data to the JobOrderState table
  4. (Optional) Configure the calendar service to record planned downtime events and shift records to time series. Refer to Use work calendars

Record machine states

Every time an equipment changes state, it is persisted to the time-series table EquipmentState.

EquipmentState table schema

    CREATE TABLE IF NOT EXISTS EquipmentState(
  EquipmentId SYMBOL,
  ISO22400State VARCHAR, -- ADOT, AUST, ADET, APT
  time TIMESTAMP
) TIMESTAMP(time) PARTITION BY MONTH DEDUP UPSERT KEYS(time, EquipmentId);
    Note

    This table shows a QuestDB specific schema. You may also add additional columns as required.

    To use the service for another time-series DB, get in touch.

    [
 {
   "EquipmentId": "Machine A",
   "ISO22400State": "ADET",
   "PackMLState": "Held",
   "time": "2024-03-28T13:13:47.814086Z",
 }
]
    Note This record includes an additional field, PackMLState, to show that additional data can also be recorded.

    Record quantity records

    You can persist two categories of quantity records:

    1. (Optional) Values generated by the machine.
    2. Final produced quantities (these should be categorised into Good, Scrap, and Rework).

    QuantityLog table schema

      CREATE TABLE IF NOT EXISTS QuantityLog(
  EquipmentId SYMBOL,
  Origin SYMBOL, -- Machine, User
  QtyType SYMBOL, -- Delta, RunningTotal (running total not currently supported)
  ProductionType SYMBOL, -- Good, Scrap, Rework
  Qty FLOAT,
  time TIMESTAMP
) TIMESTAMP(time) PARTITION BY MONTH DEDUP UPSERT KEYS(time, EquipmentId, Origin, QtyType, ProductionType);

      [
 {
   "EquipmentId": "Machine A",
   "Origin": "Machine",
   "QtyType": "Delta",
   "ProductionType": "Unknown",
   "Qty": 6,
   "time": "2024-03-28T09:30:34.000325Z"
 }
]

      [
 {
   "EquipmentId": "Machine A",
   "Origin": "User",
   "QtyType": "Delta",
   "ProductionType": "Good",
   "Qty": 10,
   "time": "2024-03-28T09:30:34.000325Z"
 },
{
   "EquipmentId": "Machine A",
   "Origin": "User",
   "QtyType": "Delta",
   "ProductionType": "Scrap",
   "Qty": 2,
   "time": "2024-03-28T09:30:34.000325Z"
 },
{
   "EquipmentId": "Machine A",
   "Origin": "User",
   "QtyType": "Delta",
   "ProductionType": "Rework",
   "Qty": 1,
   "time": "2024-03-28T09:30:34.000325Z"
 }
]

      Record job response records

      Job response records persist to JobOrderState and are used to identify the current planned cycle time of each part produced from the machine. When an operation starts, a record is created setting the planned cycle time. When the operation is finished, another record is created to reset the planned cycle time to 0.

      JobOrderState table schema

        CREATE TABLE IF NOT EXISTS JobOrderState(
  EquipmentId SYMBOL,
  JobOrderId SYMBOL,
  PlanningCycleTime FLOAT, -- Number of seconds per produced part
  time TIMESTAMP
) TIMESTAMP(time) PARTITION BY MONTH DEDUP UPSERT KEYS(time, EquipmentId, JobOrderId);

        [
 {
   "EquipmentId": "Machine A",
   "JobOrderId": "Order001",
   "PlanningCycleTime": 100, 
   "time": "2024-04-02T14:32:21.947000Z"
 }
]

        [
 {
   "EquipmentId": "Machine A",
   "JobOrderId": "Order001",
   "PlanningCycleTime": 0, 
   "time": "2024-04-02T14:59:58.947000Z"
 }
]