Additive Data Context

You have thousands of data points in your organization coming from different layers of the automation stack.

You are trying to turn the data into information by adding several other data sources and logic.

• The proximity sensor generates 0 or 1 with a timestamp.
• PLC has the position data for the sensor, identifying it is the sensor for the lid of the reactor to detect if it is open or not where 0 is closed and 1 is open, sensor name is MH250123

From this point on every layer has a logic element so you might say we can push everything to PLC so I can calculate the mean time between failures (MTBF) here. Well, that will be total nonsense, you don’t put the cart before the horse.

• SCADA generates an alarm because the lid should be closed since it is filling and it is a safety critical problem with a High-High level alarm but the interlocks are not shut down.

Now this is where it gets interesting because the PLC did not know the filling was on when the sensor generated value “1” You might say ok but we can push that to PLC… You know the answer to that from above, yes you can in some cases but why not use a system specifically designed for these?

• Digital Maintenance Management System (DMMS) which is also widely named as Computerized Maintenance Management System (CMMS) generated a maintenance work order categorized as “Instrumentation” to replace the sensor. Since there isn’t a safety risk, the raised alarm is concluded as sensor malfunction, and the time the work order opened is used to calculate MTBF for the reactor.
• Since a sensor is consumed in the work order, DMMS, or sometimes this is centrally managed by the warehouse management system created a purchase request in ERP by checking the stock level for the sensor which is now below the threshold.
• The lead time of the sensor from the last PO increased from 3 weeks to 7 weeks for site A according to the latest quotation from supplier X. Procurement analyzed they need to issue the PO for at least 5 sensors to minimize the costs. On the other hand, there is supplier Y with a lead time of 8 weeks for the sensor where procuring 3 sensors is still better in terms of unit costs including shipment.
• The central procurement department through its BI tool visually identified the frequency to consume the sensor for site A is on average 3 weeks and the current stock level is 2. There is a risk of stock-out!

You might ask why the lead time change is not reflected to update the threshold. This is where tapping into a digitally connected supply chain comes in place.

• ?AI team analyzes to predict if any other spare parts have stock-out risks. If considering mean delivery times, it seems OK however based on the historical performance of lead time compliance 3 spare parts are identified as risky with %98 confidence, and those parts’ PO are issued immediately.

?If you are linearly integrating every layer of these then you are doomed. This is one of the reasons why I love UNS so much, because everything, everyone can access real-time information through a common infrastructure.?

ISA-95 Hierarchy of the UNS in a JSON format will be like this:

Enterprise Chem Co.

?????????Site A

????????????“ERP”: {

??????????????Spare Parts: {

“bomID”: “MH250123”,

??????????????????“Lead Time of Suppliers”: ?[

{“SupplierID”: “X”, “duration”: “7 weeks”},

{“SupplierID”: “Y”, “duration”: “8 weeks”},

]

?????????????},

??????????????Minimum Number of Units for Cost Optimized PO: {

“bomID”: “MH250123”,

??????????????????“EOQ for Suppliers”: ?[

{“SupplierID”: “X”, “qty”: “5”},

{“SupplierID”: “Y”, “qty”: “3”},

]

???????????},

?????????????“BI Tools”: {

????????????????“Spare Part Consumption Period”: [

{“bomID”: “MH250123”, “average”: “3 weeks”, “Stock”: “2”},

{“bomID”: “SparePart1”, “average”: “4 weeks”, “Stock”: “2”},

]

????????????},

???????????????????Area 51

??????????????????????Line Perf

????????????????????????Reactor 1

?????????????????????????????“Sensor”: {

???????????????????????????????????????“value”: “1”,

???????????????????????????????????????“timestamp”: “1688681337036”

?????????????????????????????},

?????????????????????????????“PLC”: {

???????????????????????????????????????“sensorID”: “MH250123”,

???????????????????????????????????????“sensor type”: “Proximity”,

?????????????????????????????????????? “sensorpos”: “Lid1”,

???????????????????????????????????????“rawLow”: “0”,

???????????????????????????????????????“rawHigh”: “1”,

???????????????????????????????????????“status”: “Open”

?????????????????????????????},

?????????????????????????????“SCADA”: {

???????????????????????????????????????“Filling status”: “Running”,

???????????????????????????????????????“Interlock status”: “Open”,

?????????????????????????????????????? “Alarm Level”: “H-H”

????????????????????????????},

?????????????????????????????“CMMS”: {

???????????????????????????????????????“Work Order Number”: “2023-07-07TW1”,

"Work Order Status": "Open"

???????????????????????????????????????“Work Order Category”: “Instrumentation”,

?????????????????????????????????????? “Required Parts”: “MH250123”,

???????????????????????????????????????“MTBF”: “400 hours”

?????????????????????????????},

?????????????????????????????“ERP”: {

???????????????????????????????????????“Purchase Request”: “2023-07-07TPR1”,

?????????????????????????????????????? “Required Parts”: “MH250123”

?????????????????????????????},

Now any system can access these and consume/publish to this infrastructure. It still boggles my mind how come all of the manufacturers are not using this architecture.

Let’s schedule a call to discuss how we can help you with MaestroHub to skyrocket your digital transformation.