Machine Integration Notes

Introduction

The session focuses on machine integration with MES (Manufacturing Execution System).
The discussion involves Costa, Lego, and Nico, with Paragiotis joining later.

Blueprints and Automation

Costa shares his screen, displaying ARW HCI (presumably a system interface) with a list of machines and automation information.
The machine list includes boiler, vertical, milling, coils, forming, welding, scaffolding, etc.
Status is mandatory; if a coil is missing, the line must stop, controlled by PLC (Programmable Logic Controller) without MES.
Process control types are not for MES.
The current state is based upon blueprint, to determine if stations need to be added and how many.
There are six PLCs for employer, with interest in accumulators, forming, welding, scaffing, and annealing.
The numbers represent OPC (OLE for Process Control) tags.
Every machine has a SCADA (Supervisory Control and Data Acquisition) system with real-time reporting.
Applications used in automation include QPMS (Quality Production Management System).
QPMS handles Wick settings and pipe ID generation and marking.
QPMS reads initial visual inspection data, functioning as an MES station after the one decline.
The system checks width, cut loop, milling, welding, and ultrasonic tests, marking the pipe/coil and giving information to the operator.
Alerts are given to the operator if findings are out of limits.
QC9 is voice control in-line process, while cross control is process control based on sensors connected to the machine, not the product.

MES Integration and Parameters

Parameters in MES need to be specified.
The MES should read coil width, length, and thickness from PLC tags.
Scrap automation is desired to collect the weight of scrap per machine for production optimization.
Currently, scrap is not being weighed; a system is needed to provide real-time scrap weight information by order and shift.
Another automation, not for now, involves patience and perceptions to create bugs and track production weight.

Discussion on MES Stations and QC

Discussion on whether MES for appraisal should be at the same locations or elsewhere.
Consideration of using MES for downtimes, defects (acid defects, quality defects), requiring knowledge of product or machine position.
DC stations are present.
The system needs to identify which coil is where, provided by QPMS.
In the coil room, QC checks measurements and inputs them into MES.
Automation sessions record position, load setup, and apply pressure.
Load, setup, and apply press are actual values, while position setup and roll it at 3 are measured and entered by the user including left measurements.
Each line represents another point in PLC, another tag.
The last three columns are measurements put into MES.
Status column indicates whether a parameter is actual (existing) or potential (desired).
Measurement frequency (every second, hour, order) requires business automation input.

Sensor Data and Sampling Rates

The automation column represents a collection of tags or sensors, not a single tag.
Nikos clarifies that the automation column isn't single tags, but a collection of sensors and tags working to provide those measurements.
Signals from sensors are in millisecond frames, but such a high sampling rate isn't necessary for MES.
PLC collects data in milliseconds; MES frequency depends on production user needs.
Business defines the frequency of data recording in MES.
The fields in each screen of DGT (presumably a system interface) in pipe one drive what's needed in MES.
If fields are filled manually today, the aim is to fill them automatically with values from PLCs, requiring definition.
It's unnecessary to replicate HMI (Human-Machine Interface) functions in MES.
Alarms in the station don't necessarily need to be mirrored in MES.
The blueprint captures factory events, with automation providing data.
Information for historians server is separate and used for MES.
Determine what is exactly for MES and what ERP (Enterprise Resource Planning) needs.
Last year's effort to capture machine data can be used if needed; otherwise, it can be kept in an insurance center.
For appraisal, focus on specific tags needed, and a workshop will rearrange and define these tags.

Data Collection and Communication Methods

Appraisal can react to every change in PLC values, avoiding database overload.
Data is read and recorded only upon change (e.g., from 75 to 77.751).
This minimizes unnecessary data logging.
Another method involves sending parameters to a robot (e.g., KUKA), switching a flag, making measurements, and reading values at once.
The actual action (field for information or alarm) depends on business requirements.
Alarms can be raised for operators if values are outside limits (e.g., three sigma).
This can automatically save as a defect or nonconformity.
Alerts can be sent to supervisors in a control room, with visual indicators for machine status.
For lab robots, data is saved as measurements for reporting and certificates, stored in the database, and used for graph reporting.
Historian is a better place for such reporting.
Communication between MES and values is via OPC UA.
OPC UA server collects values from PLCs, communicating with MES through a machine integrator.
Including a tag in the system involves generating an alarm, changing a color, or storing it in the database for reporting.
Additional communication involves controlling conveyor movement or giving commands to the machine via PLC.
Emergency signals should be handled through PLC, not MES.
Distinguish between what's seen in PLC systems (HMI, SCADA) and what's in MES, avoiding a one-to-one correspondence.
Critical data for MES and analysis should be taken from automation via machine integrator or other databases (e.g., QPMS).

Defining Scope and Business Needs

Focus only on sensors marked with "yes" in the QC in process call loop.
Specify how to utilize each tag.
Transfer them via UA to the MES database even if they are not used in the screens just to have them?
A lot of interesting things were mentioned that would be of use for production people.
Whenever a process value goes out of range, might trigger a decision, and the information would be passed on to the operator.
The business SME needs to decide on decision functions or nonconformity functions.

Pipe Location and Sensor Identification

Identify tags that indicate pipe location.
Sensors in the first row of the Excel are for the coil, not pipes.
The excel you are currently sharing is not the unit of pipe.
An actual unit of pipe is created after the PTA show.
There are sensors between production stages that give information about the movement of a pipe.
If tags are missing, they can be included in the OPC server.
The need, frequency, and information from sensors in QC in process with one need to be specified.
Some statuses are "potential," meaning no capturing at the moment.
For example, weight is measured at the beginning of production.

Process Optimization and Alerts

Theoretical pipe concept: if the pipe is not in the expected location, trigger an alert.
When the coil is on the edge, a signal is sent to cut it from the PTA, generating a signal and getting the corresponding signal to the QPMS.
This relates the signal to the fiber being used from the coil.
For Alex, millisecond-level data is possible.
But the operator should not do actions that is based on automations.
Work order and time need to be consider on on the screens to create screens needed after from what we already have.
The data collected is how things work today, which we what we know, cost being the cost of having more knowledge to better our processes.

Defining Business Use Cases

Need to be more specific on what is needed in MES and the two potentials in MES, control and report.
Sensors are installed to improve automate the current manually capture of data, allowing team to focus on other things and reduce errors.
The business and production team needs to provide the data that is need from sensors.
Preventing maintenance need to be considered when parameters like power in motors, amperage in motors, etc. go out of range.
Need need to trigger alarms and Aviso will trigger an action based on that event the action.
Create defect or notify it into SAP.
We are going to create a conceptual design that will define the scope based on which we will create the detail project plan, which will be finalized.

Types of Communication and Business Cases

Business should identify examples of how data will be used.
One communication is alarm when sensors reads when the sensor is out of the range.
For the business perspective, look closer.
Creating alarms is one business cases with data monitoring being another, some other communication being third.
Within the next two weeks define the different business types.
There are several use case across different business cases and the sensors databases, which we are looking to a public server to get after the process.
We are also going ask CPI if something is missing that is needed, we can add something. Is configure with MES.
CPW, if we miss something, to add it in the later stage.
When we are on an automation and automation for a new customer needs new fields, those will new additional inspection that needs to be captured.

ID Generation and OVD

Black box for ID generation needs investigation to ensure proper integration with MES.
OVD (presumably a defect tracking system) should be relatively easy with sensors and the approach of pipe one.
The machine stops, a status changes, and the operator puts in a reason code.
The next meeting will discuss changes and interest in the database to calculate OE (Overall Equipment Effectiveness) reliability performance.

Scada Reporting and Key Parameters

Offline machines use SCADA for reporting, with different parameters (temperature, sizing, PTA).
The point where we can send the waiting part.
We can also put the tolerance in MES.
Have formation and process it also in MES.
Decide which are relevant for.
It's used with plate feeding is a special form of communication that needs investigation.
Make this exercise again.

Prioritization and Next Steps

Prioritize each line and machine to decide if integration is needed.
Highlight relevant lines with yellow for MES integration.
The remaining can be deleted in further analysis.
Each lines will be considered to if they X Rays are required to needed with MES and if they can be deleted or not.
Each business case will need use in the design design.
Review of current data inputs the current processes and the future.
Especially that here there is in Elsao will be a interface when they use the file, which will be in appraiser.
Find the length of the weight with sensors.
For reporting side how it relates to APP and historian to create that needs to be discussed.
We are going to select that the the scope and then also whatever.
Have in the functional specification is just to at least describe it in a high level how this will work