Traditionally, CPG manufacturers have relied on either legacy machine vision techniques or human inspectors to identify quality issues during the manufacturing process. However, both of these approaches have their limitations: Legacy machine vision systems can only detect simple quality problems (such as basic gauging or missing parts) and do not have the flexibility to identify infrequent subjective defects. Human inspectors, on the other hand, can recognize a broader range of unique quality issues and help identify patterns and find root causes, but they can be prone to human error and fatigue, and definitely can’t work nearly as fast as machines do.
Both solutions are also not ideal for data analytics. Traditional machine vision systems are often site specific, assembled by local integrators, and offer no consistency in measuring and reporting across the production line. Human inspectors are even less reliable in creating accurate documentation on defects and inconsistencies they find, making it harder to analyze trends and implement improvements.
AI-DRIVEN VISION INSPECTION FOR CPG MANUFACTURING Elementary: the best of both worlds for CPG companies This is where Elementary comes in. Elementary provides a full-stack vision solution that combines the robustness and consistency of machine vision inspections with the flexibility and intelligence of human inspections. It uses machine learning and cloud connectivity to create a unique system that offers specific benefits for CPG companies.
Cosmetic defects: These defects do not necessarily affect the product's function but can impact the consumer's perception. Cosmetic defects can range from dented bottles to missing caps to any number of issues that affect how the product looks. Legacy machine vision systems cannot detect such anomalies, as they are only programmed to identify specific defects. Elementary solves this problem by using machine learning-powered anomaly detection. Elementary’s vision systems can learn what an ideal product should look like and are able to identify any defect that makes the product that’s coming down the production line look different from the standard. Label defects: It is essential to ensure that the correct label is on the product, and that the label is free from errors, legible, and properly placed on the container. Elementary's solution goes a step further by not only confirming the correct label but also ensuring that the label meets all the regulatory requirements.Coding and cross-packing: Each region has its regulations related to ingredients, allergens, date codes, bar codes, or the material the container is made of. Elementary’s solution helps manufacturers confirm that all this information is correct and compliant with local regulations. Root cause analysis: In the manufacturing problem, repeat defects often have an identifiable root cause, such as machine settings that may require periodic adjustments. Elementary's automated visual inspection solution can classify defects and identify the root cause , allowing manufacturers to make necessary modifications before more defects occur. Analytics : Quality process improvements require a good understanding of what’s happening on the factory floor at any point in time. Elementary does this by automatically analyzing common defects, defect rates, and quality comparison across batches, and presenting results in a clear and readable format. This is a vast improvement over manual data collection or information gathering practices used by legacy systems today. Traceability: Elementary's cloud-connected cameras provide full traceability out of the box. All inspection results and images are stored in the cloud for up to one year to allow for complete traceability of every product inspected in the event of a quality control issue or to perform historical analysis. Flexibility : Unlike legacy camera providers who use different cameras for each inspection task (one for bar codes, another for measurements, etc.), every one of Elementary's cameras can perform all types of inspections. Elementary's cameras can also be easily configured remotely , making it easier for manufacturers to repurpose equipment or add new products or inspections.Elementary solutions are unique because they don’t rely on integrators to individually acquire, install, configure, and program an inspection system for each production line. Elementary offers CPG companies a complete visual inspection system that is equally adept at simple code readings and complicated subjective inspections, provides complete traceability and root cause analysis, and is flexible to adapt as the manufacturer's needs change.
A container is a crucial component of any CPG product, and it is essential to inspect it thoroughly before sending it out to the market. But CPG manufacturers also know that one finished product can look slightly different from another, and still be a good, quality product. One of Elementary’s greatest advantages lies in using machine learning to learn that “good” can have many variations. Our technology combines the best aspects of legacy machine vision and human inspections, paired with the latest advancements in artificial intelligence and cloud computing to perform inspections that allow for desired levels of variability without opening up parameters and sacrificing quality. Elementary solutions are designed to help CPG manufacturers identify a broad range of issues, improve production line efficiency, and ultimately deliver higher quality products – and containers that they come in.
