Platforms and vendors
- Ignition
- SQL historian
- Power BI
- Excel
- Rockwell
- ISA-95
Plant Intelligence · 13
Using Production Data to Find and Fix Losses
Metromotion Controls turns the data your lines already produce into improvement work: where yield, throughput, changeover and cleaning time are being lost, why, and whether a change on the line actually fixed it.
How we approach it
Engineered for your site and support model
Most plants collect more data than they use. The historian fills up, the OEE screen shows a number, and the improvement meeting still runs on opinion. Useful analytics starts from a question the plant cares about, such as why one SKU loses more yield or why CIP on one circuit takes longer, and answers it from data that is known to be right. Metromotion Controls does this as the team that programs the lines, so the signals are checked against the logic, and fixes that need a change in the PLC or SCADA can be made by the same people.
01
Loss analysis for yield, throughput, changeover and CIP time
02
Signal validation against the PLC logic before analysis
03
Context added to data by line, SKU, order and shift
04
Analysis in Power BI, Excel or SCADA on a SQL data layer
05
Control system changes to act on the findings
06
Before and after measurement of each change
Delivery context
Relevant experience
- Chobani: OEE and production analytics supported continuous improvement work across yield, throughput, CIP duration and scheduling.
- Chobani: consistent reason codes across shifts gave the operations team a downtime Pareto they could use in daily reviews.
- Cobs Fine Foods: checkweigher and metal detector data connected with the rest of the line, so quality losses sit beside performance data.
Section 01
Start with the loss, not the dashboard
Analytics projects stall when they start with a tool and a wish to see everything. They deliver when they start with one loss the plant already knows about and a decision someone will make with the answer. We agree that question first, then work back to the data needed to answer it.
- Yield and giveaway: product lost between ingredients in and good units out.
- Throughput: slow running and short stops that never show as downtime.
- Changeover: time lost between SKUs, by SKU pair and by crew.
- Cleaning: CIP duration and frequency against what the product and plant need.
- Quality: rejects and holds traced back to the conditions that produced them.
Section 02
Data that can be trusted first
An analysis is only as good as the signals behind it. A count from the wrong tag or a state that means two things will produce a confident, wrong answer. Before analysing, we trace each signal to the PLC logic that sets it and check it against what the line did. Where context is missing, such as the SKU, order or shift, we add it in the data layer so every record can be grouped the way the question needs.
Section 03
From data to a finding
The analysis itself is usually straightforward once the data is right. Most findings come from comparing like with like and looking at the distribution, not the average.
- Pareto of losses by cause, line, SKU and shift.
- The same product compared across lines, crews and time of day.
- Measured changeover and CIP durations against the planned ones.
- Process conditions compared between good and poor runs.
- Before and after comparisons when a change is made on the line.
Section 04
Taking the answer back to the line
A finding is only useful when something changes. Sometimes that is a work practice or a schedule. Often it is in the control system: a sequence step that waits too long, a set-point that drifts, a CIP step that runs longer than validation requires, or a missing alarm. Because we program and support the control system, the fix can be made, commissioned and then measured with the same data that found the problem.
Section 05
Analytics tools and where they fit
Most analysis needs a clean SQL data layer and a tool the site already knows, such as Power BI or Excel, rather than a new platform. Live views stay in SCADA, where operators work. Specialist analytics or machine-learning tools can be added later when a question needs them, and they work better on the same structured data. For reporting surfaces, see production reports and dashboards.
Frequently Asked Questions
Common questions
What is manufacturing data analytics?
It is the use of production data from PLCs, SCADA, historians and quality systems to find where a plant loses output, yield or time, and to check whether changes fix it. In a mid-size food plant it is mostly careful comparison of trusted data, not complex models.
Do we need AI or machine learning?
Usually not to start. Most losses are found with clean data, the right context and simple comparisons. Machine learning can help with specific problems later, and it needs the same trusted, structured data to work.
What results should we expect?
That depends on the plant and the losses. We do not quote improvement figures we have not measured on your line. The first piece of work is sized so its result can be measured.
Can you work with data we already collect?
Yes. Most plants already have useful data in a historian, SCADA or spreadsheets. We check it against the line, fill the gaps in context, and start from there.
Related Work
Connected services, industries and proof
Service
Plant Intelligence: factory data and OEE
OEE, reports, quality, CIP, maintenance and energy data from the team that builds and programs the lines.
Service
OEE and production monitoring
Live OEE, line status and stop reasons read directly from the PLCs on food and beverage lines.
Service
Production reports and dashboards
Shift, daily and weekly reports generated from line data and delivered to SCADA, Power BI or email.
Service
Reliable plant data
The right tag for each count and one meaning for each state, validated against the line and placed in context.
Insight
Power BI for manufacturing
Getting PLC and SCADA data into Power BI cleanly: the SQL layer, context and refresh limits.
Insight
OEE you can defend
How availability, performance and quality are actually calculated, and why source-of-truth machine data matters.
Insight
Breaking down plant data silos
The unified namespace, MQTT and Sparkplug, OPC UA, historian contextualisation and ISA-95 alignment for a single source of plant truth.
Project
Chobani greenfield yoghurt plant
Large capital project delivery across process automation, electrical engineering and commissioning.
Project
Chobani OEE data platform
OEE and downtime reporting from PLC states, with shift and weekly management reports for an Australian yoghurt plant.
Project
Cobs OEE data platform
OEE data platform and line integration for production performance reporting.
Service
Ignition SCADA integrator Melbourne
Ignition Gold Certified System Integrator for Ignition SCADA, historian, reporting and MES functions.
Industry
Food and beverage automation Australia
Automation, traceability, CIP, SCADA and production data for Australian food and beverage plants.
Industry
FMCG automation Australia
Machine automation, line integration, OEE data and controls upgrades for high-volume manufacturing sites.
Related work
Related project proof
Project
Chobani
Chobani needed a controls partner to help establish its Australian manufacturing operation with new production lines and site infrastructure, and the relationship has had to keep adding and optimising lines without disturbing production already running on site. Metromotion Controls supported the initial setup and has continued to add, optimise and support production lines over the years, bringing process, packaging and site services onto common control and data interfaces so each addition extends one operating environment rather than a separate island. The scope spans plant services including boilers, refrigeration, trade waste, power monitoring and water usage monitoring, plus OEE, site data analytics and contextualised production data from lab equipment and other external sources.
Project
Chobani
Building on the greenfield automation delivered at Chobani's Australian facility, Metromotion Controls was engaged to develop an OEE data platform that gave the operations team structured visibility of line performance, availability losses, and production output. The project connected existing PLC infrastructure to a centralised reporting environment, establishing consistent downtime reason codes, shift-level OEE calculation, and management dashboards aligned to how the site teams already ran daily reviews.
Project
Cobs Fine Foods
Cobs Fine Foods is one of Australia's leading premium snack manufacturers. The business needed real-time production data from all line equipment, including checkweigher and metal detector systems, connected to a centralised OEE platform provided by a third-party vendor. Metromotion Controls designed and delivered the automation integration layer, connecting each line asset to Ignition, capturing data at PLC level, and passing it upstream to the OFS OEE platform. The project included full network architecture design, device configuration, and the PLC logic required for clean, consistent data capture across the line.
Related pages
Plant Intelligence
Plant Intelligence
Factory data, OEE, reports, quality, CIP, maintenance and energy from the team that builds and programs the lines.
Plant Intelligence
OEE & Production Monitoring
Live OEE, line status and stop reasons read directly from the PLCs, shown to operators, supervisors and managers when each of them needs it.
Plant Intelligence
Production Reports & Dashboards
Shift, daily and weekly production reports generated from line data and delivered to SCADA screens, Power BI or email, with no typed entries or spreadsheet collation.
Plant Intelligence
Reliable Plant Data
Accurate, contextualised plant data built from the right tags and states, validated against the line during commissioning, so reports and dashboards can be trusted.
Want to find out where the line loses time and product?
Speak directly with an engineer about scope, timing and technical constraints.