Which process data should be captured for seamless batch traceability?
For robust batch traceability, material, process, quality and logistics data must be linked in such a way that every finished-product batch can be traced back to the raw-material batches used, and vice versa. This forward and backward tracing makes it possible to narrow down affected batches specifically in the event of a deviation, rather than pre-emptively blocking entire production periods or product families.
On the material side, the supplier, supplier batch, internal batch or lot number, goods receipt, inspection and release status, storage location and the quantity actually used per production batch are required. In addition, there is the released recipe or bill-of-materials version, raw-material identification – for example via barcode, RFID (radio-frequency identification) or DataMatrix – and, for relevant materials, information on storage time, expiry date and first-in-first-out or first-expired-first-out strategies. First in, first out means that goods put into storage first are consumed first; first expired, first out prioritises materials with the earliest expiry date.
For the manufacturing process, the plant, the equipment used, the recipe version, the operators involved and the time stamps of all essential process steps should be documented. This includes charging, dosing, mixing, emptying, cleaning, product change and releases. Relevant process parameters depend on the application, for example temperature, pressure, flow, speed, torque, mixing time, dosed quantities, energy input and moisture. Equally important are target-actual deviations, alarms, downtime, manual interventions and parameter changes with user ID, time and justification.
Not every available measured variable needs to be stored permanently. What should be recorded above all is data relevant to quality, safety, process capability, root-cause analysis or regulatory requirements. This can also include ambient conditions such as temperature, humidity, hygiene or cleanroom status, as well as the software, recipe and calibration states valid during the batch period. Maintenance events are only directly relevant to the batch where they could affect the process, the measurement capability or product quality.
The quality data must establish the link between process and release. This includes in-process checks, laboratory results, sampling plans, test equipment used, release status, deviations, rework, scrap and documented measures. Statistical Process Control, abbreviated SPC, refers to the statistical monitoring of process data; figures such as process capability indices can be useful but are not required for every batch production. What matters is that the test strategy and acceptance criteria suit the product and the risk.
For traceability after manufacturing, packaging, storage and shipping data are needed. This includes packaging and labelling data, the assignment of batch to container, pallet or shipping unit, delivery date, delivery documents, recipients and, where applicable, transport conditions. In the food industry, European law generally requires that companies be able to identify the immediate supplier and the immediate customer of a product. This principle is referred to as "one step back, one step forward"; further-reaching requirements may apply depending on the product and legal framework.
The data should be brought together in an electronic batch record or an equivalent system. A manufacturing execution system, abbreviated MES, can support operational production control and batch documentation; a supervisory control and data acquisition system, abbreviated SCADA, serves to monitor and record technical processes. An enterprise resource planning system, abbreviated ERP system, manages orders, materials and logistics, among other things. What matters is not the individual system, but the unambiguous, complete and permanently traceable linkage of the data across system boundaries.
In regulated environments, the data must be protected against loss, unauthorised change and unnoticed deletion. EU-GMP Annex 11 requires, among other things, that stored data be checked for accessibility, readability and accuracy and remain available throughout the entire retention period. Audit trails and relevant metadata must be protected over the defined retention period. The specific retention period does not follow generally from batch traceability itself, but from the type of product, contractual requirements and the applicable law. A blanket period of 10 to 15 years is therefore not reliable. Nor does 21 CFR Part 11 itself establish a uniform retention period for all cases; instead it supplements the applicable specific regulations on electronic records and signatures in each case.
Summary
Complete batch traceability links raw-material and supplier data, recipe and process data, quality and release information, and packaging and shipping data via a unique batch identifier. Particularly important are the raw-material quantities actually used, time stamps of critical process steps, relevant target-actual deviations, operator interventions and the assignment of the finished-product batch to recipients or storage locations. Data scope and retention period must be defined on a risk basis and in relation to the product and applicable regulations; general long-term periods are not valid for every industry.
Industry 4.0 integration at amixon®: recipes, data connectivity and traceability
Complete batch traceability does not arise simply from a PLC control system, a barcode scanner or a connection to an ERP system. What matters is the end-to-end linkage of raw material, recipe, weighing, process control, quality data, deviations and discharge. The mixing plant supplies an important part of the process data here; for complete traceability, however, upstream systems such as goods receipt, storage, dosing and weighing, as well as downstream systems such as packaging, storage and shipping, must also be integrated.
amixon® mixers can be equipped with a programmable logic controller, abbreviated PLC. Mixing programmes can be stored in the PLC, for example mixing time, speed, dosing sequence, temperature profile, fill level and other project-specific parameters. The PLC enables reproducible execution of released process steps. However, it does not automatically document all the information required for seamless batch traceability. In particular, raw-material origin, supplier batches, inspection and release status, actual weigh-ins, packaging data and recipient information must be integrated from further systems.
A connection to the enterprise resource planning system, abbreviated ERP system, can be provided for on a project-specific basis. An ERP system supports materials management, order management and logistics, among other things. Barcode scanners can help to uniquely identify raw materials, containers, recipe versions and production batches. This allows recipe, batch and selected process parameters to be linked with one another. Robust traceability, however, presupposes that the barcode process is organisationally secured, that all materials are actually scanned, that mis-scans are detected, and that the system data remains correctly linked in time to the respective batch.
Which process data is recorded and passed on to higher-level systems is defined at amixon® on a project-specific basis, on the basis of the user requirement specification, abbreviated URS. The URS is the operator's documented requirement specification. Depending on the mixing task, this can include, for example, mixing time, speed, torque, temperature, moisture, pressure, vacuum, fill level, dosed quantities, run times, cycles, alarm messages and operator interventions. What matters here is not the volume of data stored, but its relevance to quality, safety, root-cause analysis and regulatory requirements.
Torque, temperature or moisture can provide important indications of the process course, but are not automatically direct proof of mixing quality or product quality. Their informative value must be verified for the specific recipe, the raw materials used and the respective mixer through trials, process understanding and suitable quality data. Nor does process data acquisition replace laboratory release or product-related quality testing where this is required for the product.
The transfer of data to a manufacturing execution system, abbreviated MES, a process control system or an ERP system must be described technically and organisationally within the project. An MES supports production control, batch documentation and data acquisition. For a usable batch history, batch ID, product ID, recipe version, time stamp, plant identifier, target and actual values, user actions, deviations and releases must be brought together unambiguously. Equally important is a defined approach to communication interruptions, missing data, manual additions and subsequent corrections.
Operating data such as run times, cycles, batch counts, torque profiles, motor current, temperature, vibration and fault messages can additionally be used for condition-based maintenance. This method is also referred to as condition monitoring. It can help to detect changes at the drive, bearings, seals or discharge devices earlier. Predictive maintenance additionally requires sufficient data quality, known failure patterns, traceable warning criteria and a maintenance organisation that derives concrete measures from indications. Run-time and batch data alone are not sufficient for this.
With many amixon® mixers, the mixing tool is supported above the mixing chamber. This can reduce product-contact sealing points and support maintenance accessibility. High plant availability, however, does not arise from the design alone, but from the interaction of product design, maintenance, operation, cleaning, spare-parts strategy and the evaluation of relevant operating data. Which condition data is actually meaningful should therefore be derived from the specific wear and failure risks of the application.
In regulated applications, amixon® can provide qualification-relevant documentation and support with Design Qualification, abbreviated DQ, Installation Qualification, abbreviated IQ, and Operational Qualification, abbreviated OQ. DQ documents that the plant design meets the specified requirements. IQ confirms proper installation. OQ demonstrates that the plant functions within the defined operating range. amixon® describes its assistance on request with DQ, IQ and OQ, as well as an alignment of documentation and execution with EU-GMP and 21 CFR Part 11; the integration is intended to follow the operator's validation concept from the URS through to commissioning.
This support does not relieve the operator of responsibility for data integrity, process validation and regulatory compliance. Where electronic records and electronic signatures are maintained within the scope of 21 CFR Part 11, roles and user permissions, audit trails, electronic signatures, data backup, data archiving and the long-term readability of the data must be regulated within the validated overall system. The operator must also establish which data is required for batch release, for root-cause analysis and for statutory or contractual retention.
Pilot-plant trials with the original product can help to establish critical process parameters and suitable data-capture points at an early stage. According to amixon®, more than 30 test mixers in various sizes are available in Paderborn. There, mixing time, energy input, fill level, product protection, mixing quality and cleanability, among other things, can be investigated. The results can serve as a technical basis for recipe design, sensor selection and later production control. They do not, however, replace the batch data from commercial production, nor the validation or performance acceptance test of the complete plant at the operator's site.
Manufacturing at the Paderborn site and documented quality assurance can support the technical traceability of the plant. This includes, for example, drawings, material certificates, welding documentation, test reports and component lists. This technical documentation supports qualification, maintenance, spare-parts supply and the assessment of later changes. It is, however, to be distinguished from the traceability of the manufactured product's batches. For product traceability, the operator's consistently recorded material, process, quality and logistics data are what matters above all.
According to the company, amixon® also offers maintenance, modernisation, retrofitting and spare-parts service. These services can support the long-term operability of the plant. For demonstrable batch traceability, it is additionally necessary that changes to sensors, control system, recipes, interfaces or process parameters are assessed, documented and, where required, qualified or validated in a controlled manner.