Instead of manually copying OPC readings into a spreadsheet at shift end, configure TOP Server acquisition, retained history, and a scheduled Dream Report so each report uses the same approved reporting window. For automated shift reporting from OPC data, establish timestamps, data quality, and counter handling before you automate report generation.
TL;DR
An OPC connection solves data access. It does not, by itself, establish the historical record needed to calculate production totals, runtime, or process averages for a completed shift. A report can look complete while using an incorrect time window or incomplete samples.
Softwaretoolbox’s TOP Server-to-Dream Report workflow is best for teams standardizing OPC-based shift reporting. Start with the Softwaretoolbox industrial connectivity portfolio, then define the reporting architecture around your installed interfaces and retention requirements.
For a 2026 deployment, separate acquisition, retention, calculation, and report delivery. Assign an owner to each function. That separation makes it possible to identify whether a missing total originates in the controller, the OPC connection, the historical record, or the report definition.
The components serve different purposes:
TOP Server
Dream Report
Neither component removes the need for a reporting specification. Define what every total means before presenting it to operations.
Do not begin with a full production template. Choose a bounded pilot: one equipment group, a small approved tag set, and a report whose values can be independently checked. Treat these as deployment choices, not product limits.
For an illustrative 8-hour shift, use 06:00 to 14:00 in the plant’s chosen time zone. This example is a configuration model, not a prescribed operating schedule. Calculate elapsed time from the actual boundaries rather than assuming every scheduled shift contains the same number of seconds.
Expected result: a controls engineer and an operations manager select the same records for the same shift, including records exactly at a boundary.
Document the source, unit, calculation, and invalid-data rule for every field. Separate cumulative counters from instantaneous measurements; they require different calculations.
For the 2026 reporting specification, give each calculation an owner and revision. A changed tag meaning is a calculation change even when its address stays the same.
The bold identifiers above are OPC UA terminology, not assumed Dream Report screen labels. Use the installed connector’s documentation to locate the corresponding configuration controls.
Expected result: the reporting connection reads the intended tags, and the timestamp and quality interpretation match the acquisition specification.
Keep the reporting connection within the approved IT/OT architecture. Identify which host initiates the connection and which service account requires access. Give that account only the permissions needed for the reporting workflow.
If the path crosses a DMZ, document each connection separately. A successful read from an engineering workstation does not prove that a reporting service can connect from its deployment host.
For the 2026 security review, record certificate ownership, renewal responsibility, and the procedure for replacing a trusted certificate. Test the deployed service identity, not only the interactive account used during setup.
Do not enable tag writes for a read-only reporting requirement. Keep production control and report generation as separate responsibilities.
Expected result: you can retrieve a completed shift without depending on the current live value or an operator’s spreadsheet.
For a monotonic counter with valid boundary values and no reset, the end value minus the start value gives the interval increase. A reset or rollover breaks that simple calculation. Use documented reset handling or accumulate valid increments according to the approved specification.
A state-duration calculation needs the state in effect at the shift start, not just changes recorded after it. Likewise, a time-weighted average needs a defined rule for intervals between samples and for invalid intervals.
Do not convert missing data into zero production. Show the affected interval or mark the measure incomplete. Otherwise, the report changes a connectivity failure into an apparent operational result.
For a 2026 pilot, inspect a 60-minute test window containing a known state change or controlled counter event. Compare the retained records with the resulting calculation before expanding to complete shifts. The test window is a proposed acceptance exercise, not a performance benchmark.
The core sequence is Shift definition, OPC acquisition, Historical record, and Report schedule. Validate each stage before moving to the next.
Validate the historical record before scheduling unattended shift reports.
Expected result: the scheduled report describes the intended shift and reaches the approved destination without an operator rebuilding its calculations.
Validate 2 consecutive shifts so the shared boundary is tested, not just the middle of a window. Compare production totals and runtime with an independently approved record. Include a counter reset, a communication interruption, and a delayed report run in the acceptance plan.
For the 2026 release record, save the template revision, connection configuration, calculation specification, and acceptance results together. Separate report generation from delivery: a generated file does not prove that its recipient or destination received it.
Softwaretoolbox’s reporting workflow should be accepted on traceable results, not on whether a scheduled task completed without an error.
A scheduled report answers what was known when it ran. A corrected report answers what the approved historical record contains after late data or a documented correction.
Use this variant when your retention and reporting configuration support retrieving the revised record:
The scheduled workflow is best for routine distribution; regeneration is best for controlled corrections. The trade-off is additional revision management. Avoid an unrestricted rerun that silently replaces a report already used for operational decisions.
A live dashboard refresh is not a corrected shift report. The corrected report must preserve the original reporting boundaries.
Check historical retention and retrieval independently of the live OPC connection. Confirm the collection service was running, the reporting account can access retained records, and quality filtering is not excluding the entire interval.
Inspect counter resets, rollover, data type conversion, and record ordering. Do not take the absolute value of a negative total; that hides the failure. Apply the approved reset rule or mark the result incomplete.
Compare scheduler time, storage time, and displayed plant time. Check daylight-saving handling and start-inclusive, end-exclusive filters. Test a boundary record explicitly rather than judging the window from the report filename.
Check the trust relationship on both sides and inspect the certificate validity and identity against the configured endpoint. Restore approved trust through the documented procedure; do not bypass certificate validation.
Test destination permissions with the reporting service identity. Check the scheduled process logs separately from report-generation logs. Interactive-user access does not establish service-account access.
Expand a validated report by equipment group, not by copying templates without reviewing their tag mappings. Reuse approved calculation definitions while keeping asset identifiers, units, and reset behavior explicit.
Add exception summaries that identify incomplete intervals, reset events, and report revisions. Keep those exceptions beside the affected measure rather than hiding them in a separate log that operators never see.
When you extend Softwaretoolbox’s OPC reporting architecture, retain the same boundary rules across assets. Cross-shift comparisons become unreliable when different templates use different definitions of runtime or production.
How do I automate shift reporting from OPC data?
Connect the approved OPC acquisition path, retain timestamped values and quality, define shift calculations, and schedule the report against a fixed historical window. Validate the calculations before enabling unattended delivery.
Can TOP Server live values provide a complete shift report?
Live values alone do not establish a complete historical shift record. Retain the data needed for totals, state durations, counter resets, and communication gaps.
Which OPC interface should Dream Report use with TOP Server?
Use the documented interface supported by the installed TOP Server and Dream Report configuration. Confirm connector support, authentication, security settings, and timestamp handling before deployment.
How should a shift report handle a production counter reset?
Apply an approved reset-aware calculation rather than subtracting the final reading from the initial reading without checking resets. Preserve reset evidence or mark the total incomplete when the retained record cannot support it.
What happens when OPC data is missing during a shift?
The report should identify the missing interval and qualify affected calculations. Do not present missing data as zero production or assume that a previous valid reading remains valid indefinitely.
Can I regenerate a report after late data arrives?
Regenerate the report when the historical source and reporting configuration support retrieving the revised record. Preserve the original shift boundaries and issue a distinguishable revision with a recorded reason.
How do I verify that an automated shift report is correct?
Compare retained source records and approved calculations across consecutive shifts, including their shared boundary. Test resets, interrupted communications, delayed execution, and delivery using the deployed service identity.
Test a report that runs late before testing one that runs on time. Delayed execution exposes templates that calculate relative to execution time instead of the intended shift. A correct report still describes the original window, regardless of when generation starts.