IT systems · Practical guide

Validating FreeWeigh.Net: from a sample measurement to a documented result

A FreeWeigh.Net reading can look credible even if the wrong tare was applied, the sample belongs to another batch or a filter hides unusual results. Follow the connection between measurement, inspection configuration and decision. Use supplier evidence and independent expected results to test the actual workflow.

Table of contents

At a glance

What does FreeWeigh.Net validation involve?

Demonstrate that FreeWeigh.Net is suitable for the intended quality control process. Check the association of each measurement with the correct product, batch, unit and test configuration, together with traceable evaluation and data storage. Scope tests according to process risk. Software testing, suitability of the balance and statutory assessment of fill quantities answer different questions. One correct value or an installed module does not demonstrate the whole process.

  • FreeWeigh.Net supports statistical quality control; the specific application determines GxP relevance.
  • Tara, unit, and test configuration version are part of a measurement value's context.
  • A mathematically correct report can lead to misinterpretation due to data assignment or data selection.
  • Device traceability, software validation, and legal volume assessment remain separate.

FreeWeigh.Net is SQC software: determine the installed scope.

METTLER TOLEDO describes FreeWeigh.Net as modular software for statistical quality and process control. A central application is the sample-based net fill volume control. Additional measurement parameters and quality attributes can be incorporated depending on the equipment configuration. The subject of this article are these inspection and data processes, not a complete MES nor a chromatography data system.

Start with the installed combination: software version, licensed modules, test sites, connected devices, database, and interfaces to other systems. The public module brochure 30313138 includes, among other features, Batch Handling, Release Criteria, Test Item Versioning, and XML Link. Some pages contain a version date of 2016. Therefore, this description does not replace the current documentation for your installation.

A current numerical release version is not claimed here. In particular, during an upgrade, confirm that the functions, devices, and data formats are compatible with the intended system state. The availability of a module does not automatically activate it within your license. A connectivity capability does not imply a verified integration with your specific Wagent type.

Sources: [1] [2] [3]

GxP, net weight and measurement authority to be separately categorized

A fill volume application is not solely GMP-relevant due to its product name. First, identify the decision that is based on your data. If results from an in-process control during the manufacture of human medicinal products are considered, the relevant GMP process forms the basis of the evaluation. A purely operational or business-oriented analysis serves a different purpose.

In the context of human medicinal product GMP, Annex 11 is the applicable guideline for computerised systems. The current EudraLex volume still references the January 2011 revision. Legal requirements for finished product packaging, measurement authority, and their specific applicability must be determined separately. The figures presented in this article do not support any permissible underfilling and no release of a genuine packaged product.

A module designated as 21 CFR Part 11 does not clarify the FDA's scope of applicability. The relevant recordkeeping obligations and actual electronic usage are decisive in this regard. Ensure a justified classification is established prior to defining additional controls or documentation requirements.

Sources: [4] [5] [8]

What decision must your inspection process make?

The following matrix serves as an internal reference tool. It links implemented functions to a specific technical question. Capture only functions that are actually in use and supplement with the party requiring the result. This enables visibility into whether an existing supplier certificate is sufficient or whether an operational transition remains pending.

Prioritize based on error sequence and existing controls. An apparently minor unit conversion can be more significant than a convenient diagram. Missing information is not a valid justification for low risk. First clarify which data source and decision are affected.

Own decision matrix for specific FreeWeigh.Net usage
Area to assessProcess questionSpecific need for evidence
Test site and equipment Where does the individual value originate? Device identity, transmission format, and assignment to the actual sample.
Product and Batch Does the sample belong to the correct order? Normal case as well as incorrect or missing identity with specified treatment.
Test configuration What Tara and unit applied? The version in force at entry and a traceable history of earlier values.
Local result status Which decision triggers the status? Authority, criteria, and handover; no implied equivalence to full batch release.
Report and Export Which data were selected? Alignment of population, filters, individual values, and recipient assignment.

Sources: [7] [5] [2]

Evaluate suitability and correct data processing separately

A calibrated balance can provide an accurate reading, which is then incorrectly assigned to the wrong data set. Conversely, correct data transfer does not render an inappropriate measuring device unsuitable. Chapter 3 of the EU GMP Guideline addresses appropriate measurement range and accuracy, as well as calibration and testing. These device qualifications must be suitable for their intended use.

For software validation, additionally specify whether gross or net values are transferred and at which point tare is applied. Performing tare twice is equally invalid as omitting it entirely. Document unit, decimal format, resolution, and, if applicable, conversion factors. The displayed value alone does not indicate which processing steps have already been performed within the device.

Verify the path from origin to the stored dataset using known inputs. A manually entered test value can validate a data evaluation logic but cannot substantiate an actual device connection. When both verification methods are employed, their respective scopes should be clearly defined. The evidence should later enable identification of which risk is covered by which test.

Sources: [9] [5] [3]

Change Tara and inspection configuration without losing the audit trail

The historical manufacturer brochure describes Test Item Versioning as a feature enabling the creation of new settings while retaining access to prior raw and statistical data. Verify this function on the actually deployed system. What is critical is not merely the appearance of a new version number, but which samples are assigned to which parameter settings.

A reasonable requirement defines the transition point: Does the new Tara apply to a new batch, a new test instruction, or to a specifically designated date? What happens to a sample that has already been initiated? These rules originate from your process. They must not be inferred solely from incidental observations of software behavior.

Obtain the original context. A technically necessary correction requires a traceable evaluation and assessment of its impact; it must not implicitly reinterpret prior results. Use an independently documented as-built state for comparison. A copy of the current configuration does not automatically indicate the state against which an older report was prepared.

Sources: [2] [5] [6]

Further reading:

Example: two tower stations, a traceable data pathway

This is a software validation example, not a completed product test. The intended use is: For Test Batch A, configuration V1 with tare 10.00 g; for Test Batch B, configuration V2 with 10.40 g. The software receives gross values here explicitly; tare subtraction in the preceding device is excluded. All values are specified to two decimal places. No acceptance criteria are established for a real product.

Every case begins with its documented initial state. Inputs and setpoints are defined outside the application being tested and are fixed prior to execution. For evaluation, the calculation uses gross weight minus valid tare. The table describes expected results, not any specified manufacturer functions. Error handling is defined as a local requirement prior to testing.

The reviewer compares input, stored value, batch assignment, and parameter status. In the historical case, A remains below V1. If V2 is incorrectly applied retroactively, a value of 99.80 g results instead of the expected 100.20 g. This difference of 0.40 g represents a data integrity failure in the example, not a reflection of statutory fill volume tolerances. In addition to the mean value, verify the number and identity of the samples included. The B-statistics contain exactly two values. If A under V1 is inadvertently included in the calculation, rounding to two decimal places would yield 100.27 g instead of 100.30 g.

Expected values: net = gross − assigned tare
Situation and BackgroundInputIndependently established expected resultEvidence to retain
1 · A under V1 Gross 110.20 g; Tara 10.00 g Netto 100.20 g Dataset A, V1 and both input values.
2 · B under V2 Net weight: 110.20 g; Tare: 10.40 g 99.80 g net weight Dataset B uses the latest version.
3 · Second Sample B Gross 111.20 g; Tara 10.40 g Netto 100.80 g; average weight of the two B-samples: 100.30 g Exactly two intended B-values, without A.
4 · History after change Retrieve stored sample A again Furthermore, 100.20 g under V1 Comparison with the received original document.
5 · Unknown Batch X New test entry with no valid batch assignment Do not record as a regular result if intended use is not met; refer for clarification Error report or controlled treatment; no concealed assignment to A or B.
6 · Retesting of B-Probe 2 Resend the identical test input once again No third regular measurement required; identify and address repetitions. Inventory remains at two B-samples; feedback and processing are traceable.

Sources: [5] [6] [7]

Verify interfaces with missing and duplicate data

The module documentation identifies XML link and Data I/O as potential means of data exchange. This does not yet guarantee the detection of duplicate data. For each actually implemented data transfer, establish a data model specifying: professional identifiers, values, units, parameter references, and the meaning of a feedback response.

A successful transfer first addresses a technical question. For the process, it also matters whether the recipient fully and appropriately processes the correct dataset. Test a missing mandatory reference, a disrupted connection, and a repeated entry. Prioritize determining whether a failure results in rejection, queueing, or controlled manual clarification.

If an error scenario remains unclear, the test shall not be concluded solely with a correct nominal value. Record the result, affected data, and further usage. After correction, specifically investigate the root cause of the error and plausible additional pathways involved. For manual replacement transfers, additional risks of misidentification shall also be evaluated.

Sources: [2] [5] [7]

Reports, selection, and rights as part of the result verification process are addressed.

The free-reporting description indicates that data for reports can be selectively chosen. Therefore, an overview report still does not demonstrate completeness. Define for each specific purpose which samples, time periods, configuration states, and notable values are relevant. A permissible level of detail and a complete decision justification may differ.

Check whether the handling of a conspicuously atypical test result remains evident. If data are excluded, the reviewer must provide professional justification and the original context. A visually more favorable statistical presentation is not a valid justification. Similarly, the selection of a different batch must not unobservedly substitute for the actual result.

Assign permissions to tasks: collect measurements, maintain parameters, evaluate data, and manage users. Verify both authorized and denied actions. The scope of change history and review is determined based on relevant data. The mere presence of an audit trail module does not constitute proven completeness of all required events.

Sources: [2] [6] [5]

Further reading:

Use appropriate documentation and maintain the verified status.

Request available supplier and implementation documentation and align it with your requirements. A test is considered valid if, among other factors, version, functionality, prerequisites, expected outcomes, and results are clearly and adequately traceable. Open, customer-specific mappings will be supplemented. This does not constitute a blanket acceptance of all supplier documentation nor a mandatory requirement to reperform every test.

Evaluate the impact of future changes: new scale, updated driver, different calibration mechanism, report template, or interface mapping. The rationale shall specify which existing documentation remains valid and where additional assessments are required. Training, operating instructions, and responsibilities are part of the operational handover.

A restoration test should verify related values, configuration states, and history. The mere availability of a backup file is insufficient. Establish appropriate retention and review cycles aligned with the process. Historical access must also clearly demonstrate how the previous result was obtained.

Sources: [5] [6] [7]

Preparation and Checklist for QA, Production, and IT

For a small team, early clarification of work responsibilities is essential. Production or department defines the measurement task and decision-making authority. QA identifies the relevant quality requirements. IT outlines data flows and operational procedures. The supplier provides appropriate product documentation. These contributions can converge in a shared work folder; separate document silos are not required.

The effort increases especially with different device couplings, changing parameter levels, manual intermediate paths and missing historical documents. Therefore, not only count modules. Determine how many really different decision and error paths can be detected. The following checklist helps before starting:

  • Use of data and relevant regulatory areas documented.
  • Software status, modules, devices, and database recorded.
  • Gross/net value, Tara, unit and conversion clearly determined.
  • Product, batch, and test configuration assignment defined.
  • Independent target values, as well as normal and error scenarios, have been prepared.
  • Reports filter and handling of out-of-range values clarified.
  • Access rights and subject-matter review assigned.
  • Appropriate supplier documentation and remaining gaps identified.
  • Change management, safeguards, restoration, and responsibilities defined.

Sources: [5] [7]

Project experience and collaboration

Verify tare and batch assignment against known measurements

With Production, Quality Control and IT, we examine a sample alongside product, batch, tare and test configuration. We trace how these details reach the evaluated result. Known measurements support focused tests of data assignment, changed tare values and repeat transmissions. A single application may allow a narrow scope. For multiple balances and interfaces, we first map shared data paths. Equipment suitability and any required fill quantity assessment are addressed separately with the responsible specialists.

Frequent Questions about FreeWeigh.Net and Measurement Value Verification

Is FreeWeigh.Net a MES or a chromatography data system?

The manufacturer describes FreeWeigh.Net as a statistical quality and process control with an emphasis on filling quantities. Individual batch and test plan functions do not automatically make a complete MES. Chromatographic peak integration is not part of the purpose documented here. For your project, present the actual measurement, assignment and decision processes; this is the basis for the proof.

Sources: [1] [3]

When must FreeWeigh.Net be validated in a GMP environment?

It is essential to determine whether and how the application supports a GMP-regulated activity. For example, describe whether the data is used for in-process controls or quality decisions. From this, identify relevant functions and associated risks. The mere installation, product name, or use within a pharmaceutical company does not suffice to address the scope question. Additional legal purposes are separately and distinctly evaluated.

Sources: [5] [4]

Does our scale calibration suffice for the purpose of detection?

It addresses a device-related question, not all questions pertaining to data processing. Additionally, verify that the device, sample, batch, unit, and parameter settings are correctly linked. A properly calibrated balance does not prevent double tare subtraction in a subsequent data pathway. Conversely, a correct interface test neither ensures an appropriate measurement range nor replaces the necessary device validation evidence.

Sources: [9] [5]

Do we need to perform a complete revalidation for every Tarawert?

Not every parameter change requires the same level of evaluation. Assess which calculation, assignment, history, and decision is affected. Continue to use appropriate validation evidence and specifically address any outstanding implications. For example, if the method of tare determination changes from a single value to a more complex approach, the evaluation scope may need to be broader. The documented change decision justifies the extent of evaluation.

Sources: [5] [7]

Does Test Item Versioning already provide a complete history?

The manufacturer's description mentions the preservation of previous raw and statistical data. However, for your use, you can check which parameters, values and assignments are actually reconstructable. Compare an old dataset after a controlled change with its independent original document. You can clarify whether this function is included in your license and version using appropriate documentation.

Sources: [2] [6]

May unusual samples be hidden from a report?

A targeted data view may serve a different purpose than a complete substantiation of a quality decision. For the latter, the selection and exclusions must be logically justified and traceable. Verify whether a reviewer can identify the relevant data and its context. A better mean value alone does not justify exclusion. Explicitly state the purpose and population of the report.

Sources: [2] [6]

Is the "approved" status equivalent to a medicinal product batch release?

No. The local software status assesses configured fill quantity criteria. It does not replace certification of a medicinal product batch by a Qualified Person. Define the meaning of the status and any further decisions required. Where a system records certification and batch release, Annex 11 section 15 calls for restricting certification to Qualified Persons and identifying the person releasing or certifying the batch using an electronic signature.

Sources: [2] [5]

How do we verify a repeated XML transfer?

First, define how an identically qualified sample can be recognized and specify how duplicates should be handled. Then, resubmit a previously processed test entry and compare the response, data inventory, and statistics. The test must demonstrate that the agreed-upon processing procedure functions as intended. The term "XML Link" alone does not guarantee duplicate detection or a specific error response.

Sources: [2] [5]

Does the example demonstrate an allowable net fill quantity of 99.80 g?

No. The value arises exclusively from the example calculation in this article: 110.20 g minus 10.40 g. The example contains neither a legal tolerance nor a product specification. It verifies whether the intended tare is applied and the context is preserved. For actual packaging, the applicable requirements and professional acceptance criteria must be independently defined.

Sources: [1] [5]

What does a small and medium-sized enterprise need prior to conducting a reliable cost estimation?

Useful are system status, module list, device and interface overview, relevant test protocols, and an example report with associated individual values. Supplement existing supplier documentation and known gaps in evidence. From this, the scope of different testing approaches can be determined. A fixed duration derived solely from the product name would be largely inconclusive; what matters most are configuration, usage, and available collaboration.

Sources: [5] [7]

Sources and review date

As of 6 October 2026. Product description and historical module documentation from METTLER TOLEDO constitute the manufacturer's reference; a current numerical FreeWeigh.Net version is not asserted. Regulatory classification: Human medicinal product-GMP; FDA reference applies only within the relevant scope. Detailed evaluation of finished packaging and weighing regulations is not performed. Tables and target values are internal work aids and do not represent product test results. Targeted source cross-check on 18 October 2026: Annex 11, Item 15 relating to electronically recorded certification and batch release.

  1. FreeWeigh.Net — Software for Statistical Quality Control
  2. FreeWeigh.Net — Software Modules
  3. FreeWeigh.Net Datasheet — Multifaceted Quality Control
  4. European Commission: EudraLex Volume 4
  5. EU GMP Annex 11: Computerised Systems
  6. EU GMP Chapter 4: Documentation
  7. EMA: ICH Q9(R1), Step 5, Corr.2
  8. FDA: Part 11 — Scope and Application
  9. European Commission: EU GMP Chapter 3 – Premises and Equipment

Sources checked: .

Daniel Herrmann, Daniel Herrmann Consulting

Daniel Herrmann Consulting

Verify the actual measurement and decision pathway

We support you in distinguishing hardware and software scopes, evaluating existing evidence, and developing pending assessments for your quality control process.

  • Computer system validation →

    Bring your GxP systems into validated operation, from laboratory software, MES and eQMS to ERP. We define the scope, plan risk-based tests using GAMP 5 and hand the evidence to QA for release.

Send a project enquiry Customer testimonials →