FAQ¶

This page answers common questions users ask when getting started with Harro Digital.

If you cannot find an answer, contact support: harro-digital@hoefliger.de

General questions¶

What is Harro Digital?¶

Harro Digital is an on-premise platform that connects machines, sensors, and operators.

It can:

monitor real-time production data
detect inefficiencies and recurring issues
provide AI-supported guidance (depending on your deployment)

How is Harro Digital different from typical analytics dashboards?¶

Compared to dashboards that only visualize data, Harro Digital typically combines:

real-time production data
analysis and diagnostics for messages, downtime, and scrap
actionable guidance directly in the UI (if enabled)
on-premise data control (no external cloud required)

Accounts and permissions¶

Do I need a user account to use the app?¶

Often, no. Many deployments allow read-only access to dashboards and basic views without signing in.

However, a login is typically required for changes that modify the system, such as:

editing machines and machine status settings
managing users and roles
creating or editing instructions

If you are unsure what applies in your environment, ask your administrator.

Open Harro Digital.
Select Sign In (location can vary).
Enter your assigned username and password.

If additional verification is configured by your organization, follow the on-screen prompts.

How can I change the interface language?¶

Use the language selector in the header (often top-right) and choose your preferred language (if available).

Using the application¶

Why do I see a browser warning that the connection is not secure?¶

In on-premise deployments, Harro Digital may use a self-signed SSL certificate. Browsers cannot verify self-signed certificates automatically, so they show a security warning.

How can I reduce or remove the browser security warning?¶

Install the SSL certificate from your official app instance on your device (follow IT policy):

Settings > SSL Certificate Settings

See: SSL Certificate Settings

How do I switch between machines?¶

Use the Machine selector (often in the dashboard header) and select the machine or line you want to review.

See: Dashboard

What does the time range selector do?¶

The time range selector defines the time window used for most widgets and analyses (for example: "Last hour", "Today", "Last 24 hours").

Note: some live widgets can ignore the selected time range and always show real-time data.

See: Dashboard

How can I analyze a specific warning or error?¶

From a chart or message list widget:

Select the warning/error message entry.
Choose Analyze message to review occurrence history and durations.
Choose Get help to view guidance based on linked instructions (if available).

See: Analyze Message Popup

Why do OEE or performance KPIs look unrealistic (for example "3000%")?¶

A common cause is incorrect machine status configuration, especially a producing state that is missing the Running flag.

Fix:

Review status settings for the affected machine.
Correct the Running and Planned downtime toggles for relevant states.
Re-check the KPI widgets for the same time range.

See: Status Settings

Why do the parts counter and the parts-over-time total differ?¶

The Good Parts / Bad Parts counter and the total under Produced Parts Over Time can show slightly different numbers for the same machine and time range. Both panels read the same underlying production data — nothing is missing. The difference comes from how each panel defines its time window:

The Good/Bad Parts counter sums production exactly within the window you selected (for example 10:08–11:08). This is the precise figure for that range.
The Produced Parts Over Time panel groups data into fixed time buckets (for example 5-minute buckets). So the chart does not start on a cut-off half-bucket, its first bucket is extended back to its natural boundary (for example 10:05 instead of 10:08). The total under the chart then sums all visible buckets, including those few extra minutes before your selected start.

That leading slice is up to about one bucket width (for example ~5 minutes) of extra production, which is why the time-series total can read a little higher — proportionally on both Good and Bad Parts. It scales with your production rate, so at higher output it looks like a larger gap.

Which value is authoritative? For the exact range you selected, use the Good/Bad Parts counter. Expect the Produced Parts Over Time total to be equal or slightly higher depending on the panel's bucket size. To compare like-for-like, check the two numeric totals (counter vs. the total below the chart), not the chart shape.

See: Good Parts · Produced Parts Over Time

I pressed Refresh but the Good Parts counter did not change¶

This is usually expected behaviour, not a defect. The Good Parts / Bad Parts counter is a processed (aggregated) metric, not a live signal:

Production data flows through a short processing pipeline (machine/OPC UA data → ingestion → per-minute aggregation) before it reaches the counter.
These aggregates are produced in cycles and a given minute is only counted once it is complete and processed. In practice a newly produced part appears after roughly 1–3 minutes, and the counter advances in about 1-minute steps rather than continuously.

Pressing Refresh does request fresh data from the server every time (there is no cached value). If the number does not change, it is because the server has no newer aggregated value yet — not because the button failed.

This is also why the Live Status panel can show "running" while the counter has not moved: Live Status reads a real-time machine signal, whereas the counter reads the processed aggregates, which lag by the short processing delay. Both are correct.

How to confirm it is working: leave the dashboard on the machine for 5–10 minutes (or wait for the automatic 5-minute refresh) while it is running, and check that the counter steps up. If it stays completely flat for many minutes while the machine is clearly producing good parts, contact support with the machine name and the approximate time window so the processing status can be checked.

See: Good Parts · Refresh logic