Label Generation Automation System

Background

Labels were produced using an Excel spreadsheet that stored part numbers, product descriptions, and country of origin information. This spreadsheet had to be manually maintained and updated over time.

In practice, this approach caused several issues. If a part number did not exist in the spreadsheet, the person creating the label would manually enter the details. Because this relied on individual input, descriptions were not always entered consistently, and key information could be omitted — such as whether a component was for the left or right side, or the correct country of origin. This led to inaccuracies in printed labels.

Additionally, the spreadsheet was edited by multiple people across different departments, making it difficult to ensure data accuracy. Updates were typically reviewed only once per year, meaning errors could persist for long periods without being corrected.

To address this, I aimed to centralise the process by integrating the label creation workflow with the commercial system, which already stored accurate and up-to-date product data. The goal was to eliminate reliance on the spreadsheet and ensure labels were generated using a single, reliable source of information.

Project Objectives

The aim of this project was to improve both the accuracy and efficiency of label creation.

Previously, users were required to manually enter all label details, which increased the likelihood of inconsistencies and errors. To address this, the process was redesigned so that users only needed to input a part number. The system would then retrieve all associated information — including descriptions and country of origin — directly from a commercial system.

This approach ensured that labels were generated using accurate and up-to-date data from a single central source, removing the need for manual data entry and reducing reliance on outdated spreadsheets.

Error handling was also implemented to improve reliability. If an issue was identified — such as missing or invalid data — the system would display a clear error message explaining the problem and, where necessary, prompt the user to escalate the issue for resolution.

Overall, the solution significantly improved efficiency by streamlining the label creation process, while also increasing data accuracy and consistency across all labels.

Solution Design

The commercial system included a built-in reporting system that could be used to extract the required data. However, it was not particularly intuitive and required formal training to use effectively. After completing this training, I was able to design and generate reports that retrieved the correct part information needed for label creation.

This process required learning a completely new system, which initially took time. However, once established, the reports could be generated quickly and reliably. I also focused on optimising the reports, as they would be executed multiple times throughout the day. This was important to reduce load on the system while still ensuring all required data — including descriptions and country of origin — was returned accurately.

With the reporting layer in place, I then integrated the data into the label printing software. The existing label templates were adapted to support a more structured and user-friendly input process.

A new interface was designed using a carousel-style layout, allowing users to select from the available label formats. A search function was also implemented at the top of the interface. Users could enter a part number from their paperwork, which would trigger an indexed search to locate the correct record.

Once selected, the user could confirm the part, choose the required quantity, and generate the label. All relevant information was then automatically populated, including barcode data, ensuring the labels were accurate, consistent, and easily scannable.

This removed the need for manual data entry entirely and standardised label generation across all users.

Key Features

Automated Data Extraction

Label data is retrieved directly from the commercial system, ensuring only the required information is used while excluding unnecessary fields.

Centralised Data Management

All label data is sourced from a single location. If an error is identified, it can be corrected once in the commercial system, with changes automatically reflected across all labels.

Improved Efficiency

Labels are generated using pre-populated data, removing the need for manual data entry and significantly speeding up the creation process.

Controlled User Input with Automated Data Population

Users are required to search for and select the correct part number; however, once selected, all label data is automatically populated from the commercial system. This removes reliance on manual typing while still allowing user control over label selection.

Scheduled Automation

Data retrieval processes are designed to run on a schedule, ensuring the system always has access to up-to-date information.

Workflow Preview

The workflow is designed to keep user input minimal. The user enters the part number and quantity, the system retrieves the matching product details, and the final label is generated ready for printing or review.

Process Flow

Outcome

The new solution significantly improved both the efficiency and accuracy of label generation.

Previously, labels often contained spelling mistakes or missing key information due to manual data entry and inconsistent spreadsheet updates. By sourcing all data from a single, central system, these issues were eliminated, ensuring labels were generated with accurate and consistent information every time.

The process also became much faster. Users no longer needed to manually enter or verify large amounts of data, allowing labels to be created quickly and with greater confidence in their accuracy.

From a business perspective, this led to improved operational efficiency. Stock could be labelled more quickly and reliably, reducing delays in workflows. It also minimised issues caused by incorrect part numbers or descriptions, which previously could result in stock needing to be relabelled.

Overall, the solution delivered a more streamlined, reliable, and scalable process for label generation.

Testing & Validation

A significant amount of time was spent adapting each label template to support the new system and ensuring correct data mapping from the database. This included validating that each field — such as part number, description, and country of origin — was correctly populated in the appropriate section of the label.

Error handling was implemented to improve reliability. If data could not be retrieved or was missing, the system would notify the user and prevent the label from being generated. This ensured that incomplete or incorrect labels were not produced.

Additional validation was introduced to handle formatting issues. For example, if a product description exceeded the available space on the label, the system would automatically reduce the font size to ensure the text fit within the label boundaries. This maintained readability while preventing layout issues during printing.

Testing was carried out across a range of scenarios to ensure consistent behaviour, including missing data, incorrect inputs, and varying label formats. This ensured the system was robust and reliable in day-to-day use.

Current Status & Future Improvements

The solution has been in use for eight months. During the initial rollout, time was spent supporting users as they adapted to the new process. Initially, some users continued to manually verify data against the old system; however, confidence in the new approach quickly grew as accuracy improved. As a result, users began to trust the system, which further increased the speed of label generation.

One unexpected outcome was that data-related issues were often reported directly to me, rather than to the team responsible for maintaining the source data. This highlighted the importance of clearly defining ownership. Users are now being directed to the appropriate team to ensure that data corrections are made at the source, allowing updates to flow through the system automatically.

Another insight gained during early use was that some users were selecting incorrect label types for certain parts. This resulted in part numbers not appearing in searches, as the criteria for that label type were not met. Through guidance and training, this issue was resolved, and users now have a better understanding of selecting the correct label format.

Looking ahead, one key area for improvement is the ability to measure the impact of the solution. While it is clear that the process is faster and more accurate, there is currently no quantitative data to demonstrate this. Implementing metrics — such as time saved per label or reduction in errors — would provide a clearer view of the efficiency and accuracy gains achieved by the project.

Challenges & Lessons Learned

One of the key lessons from this project was how to communicate technical solutions to different audiences. Initially, I focused on explaining the technical detail to senior managers, which often led to confusion and made it harder for them to see the value of the project. I adjusted my approach to focus on outcomes — clearly explaining the benefits, expected improvements, and overall impact — rather than how the solution worked behind the scenes. This proved far more effective.

Another challenge arose after the initial rollout relating to how frequently certain data needed to be updated. While most fields — such as part numbers and product descriptions — could be refreshed periodically without issue, the country of origin required much more frequent updates due to changes in suppliers.

The original design relied on scheduled report execution, but increasing the frequency of these reports would have placed excessive load on the system. To avoid this, I temporarily reverted to the previous method while investigating a more efficient solution.

I then identified that the label generation software supported Visual Basic scripting. This allowed me to implement a targeted solution where the country of origin field could be retrieved dynamically at the point of label generation, ensuring it was always up to date (within seconds) without increasing the frequency of report execution.

After updating the label templates to support this change, the solution was reintroduced into production. This resolved the issue without impacting system performance, and no further problems have been reported since implementation.

This highlighted the importance of fully understanding data requirements early in the design phase. While the final solution was effective, identifying this requirement earlier would have avoided the need for rework.

Final Reflection

This project demonstrated the value of replacing manual, inconsistent processes with a centralised and automated solution. By integrating label generation with the commercial system, I was able to significantly improve data accuracy, reduce manual input, and streamline the overall workflow.

Beyond the technical implementation, the project also highlighted the importance of user adoption, clear ownership of data, and effective communication with different audiences. Ensuring users trusted the system was just as important as building it.

Overall, this project reflects my approach to problem-solving — identifying inefficiencies, designing practical solutions, and continuously refining them based on real-world use.