Blog

How to Connect Factory Machines and Track Performance

Learn how shop floor digitalization connects machine data, tracks downtime and OEE, reveals recurring losses, and supports phased improvement.

By ITHENA ITHENA Insights 5 min read

TL;DR

Shop floor digitalization connects machine and production data to the people who need to understand performance and act on problems. A practical system gathers signals from PLCs, sensors, HMIs, and existing records, gives them consistent context, and presents useful views of production, downtime, and quality. Start with one line, agree on what counts as a stop or loss, validate the data with operators, and use the results to investigate recurring constraints before expanding.

Key Takeaways

  1. Connect the data sources that answer a defined production question, not every signal at once.
  2. Standardize machine states, downtime reasons, and shift or product context before comparing performance.
  3. Use dashboards to find recurring losses, then verify root causes with people close to the process.

Quick Q&A

What is shop floor digitalization?

It is connecting production information and workflows so teams can see current and historical performance, identify issues, and coordinate action. The scope may include machine data, operator entries, quality records, maintenance information, and production targets.

How do factory machines send data to dashboards?

Data may come from PLCs, sensors, HMIs, databases, or production systems. A connectivity layer collects and maps the relevant signals, adds context such as line, product, shift, and time, and passes them to an application or dashboard. The exact architecture depends on the equipment and security requirements.

What should a factory monitor first?

Start with a specific problem, such as unplanned stops on a constrained line, slow cycles, or recurring quality loss. Select a small set of measures and operational context that can help explain that problem, then validate whether the data reflects what operators observe.

Stage Work to complete Evidence to review
Connect Identify machines, available signals, protocols, ownership, and network constraints. Signal quality, timestamp consistency, and reliable data capture.
Define Agree on production states, downtime reasons, planned time, ideal cycle, and quality counts. Operators and supervisors interpret the measures consistently.
Visualize Build role based views for current status, trends, and losses. Teams can find a meaningful exception without searching through unrelated metrics.
Improve Investigate recurring losses, assign actions, and review whether they were effective. Actions have an owner, due date, and follow up measure.

Industrial IoT Standard Operating Procedure

1. Trace data from the machine to the decision

For each use case, map the path from source to action. A PLC or sensor produces a signal; a gateway or connectivity service reads it; an application associates it with a machine and production context; and a dashboard or alert presents it to a role. Decide how the data will be time-stamped, retained, and protected, and how the system behaves if a connection drops.

Not every machine needs the same integration. Some equipment already exposes useful data through a standard interface. Other assets may require a sensor, a manual input, or a limited data collection approach. Involve controls, IT, operations, and security stakeholders before connecting production equipment. ITHENA describes iMEX as collecting data from machines, databases, HMIs, and systems of record, with connectivity options including OPC UA. Review the published iMEX manufacturing excellence overview and Industrial IoT solution for product context.

2. Make the measures consistent.

Machine data is only comparable when the underlying definitions are clear. Decide what constitutes planned production time, a running state, a stop, a minor interruption, and a reject. Define how product changes, breaks, setup, and shift handovers are represented. Keep a controlled list of downtime reasons that is detailed enough to guide action but simple enough for the team to use accurately.

OEE is commonly expressed as availability multiplied by performance multiplied by quality. To calculate it meaningfully, teams need agreed definitions for the time base, ideal cycle, and good output. A dashboard can display a calculated OEE value, but the number is not useful if machines classify stops differently or production context is missing. Treat the formula and its inputs as a shared measurement standard, not a black box.

3. Turn visibility into loss analysis

Review performance by line, machine, product, shift, and time period where the data supports those comparisons. Look for recurring patterns: repeated short stops, a machine that limits the line, extended changeovers, or a rise in rejects after a particular condition. Use the trend to focus an investigation, not to declare a cause automatically.

Validate the pattern with operators and maintenance staff. Check whether the event coding is accurate, compare it with production records, and observe the process where appropriate. Assign an action to an owner and verify whether the loss changes after the intervention. This closes the loop between production data collection and continuous improvement.

4. Design views for each role

A production supervisor may need current line status, target versus actual output, and open issues. A maintenance lead may focus on downtime events, repeat failures, and response status. A plant manager may need a consistent view across areas and trends in output, availability, quality, and delivery. The same underlying definitions should support these views, while each screen emphasizes decisions relevant to its audience.

Digital workflows can complement machine monitoring. ITHENA's iGEMBA digital shop floor describes digital Gemba checklists, KPI views, issue tracking, and corrective action workflows. This can help teams connect performance signals with daily review and follow-up.

5. Roll out one line at a time.

Select a line with a clear operational question and a willing cross-functional team. Inventory its equipment and signals, confirm network and security requirements, define the measures, and test the data against observation. Train users on the dashboard and downtime codes, then run the system long enough to expose gaps in definitions and data quality. Review actions and outcomes before extending the approach to another line.

Conclusion

Shop floor digitalization creates value when machine data is trustworthy, contextualized, and connected to an improvement routine. Begin with one line and one question, map data from source to decision, establish shared metric definitions, and use trends to guide investigation. Expand in phases as the team builds confidence in the data and the actions it supports.

Sources

  1. ITHENA Manufacturing Excellence iMEX: published description of data collection, connectivity, dashboards, alerts, and manufacturing use cases.
  2. ITHENA Industrial IoT: product destination supplied for Industrial IoT; automated page verification was unavailable during drafting.
  3. ITHENA Digital Shopfloor iGEMBA: published description of digital Gemba, KPI views, issue tracking, and follow-up.

Next Step

Choose one line and define the production loss you want the data to help explain. Explore ITHENA manufacturing visibility solutions.

Get started

Connect one line. See the deviation before it becomes downtime.

Start with one line on your own data and expand use case by use case.