Consolidating In-Line Bottle Inspection with AI Edge Inference Computers

Premio Case Study featuring AI Edge Inference Computer, RCO-6000-CML, for in-line bottle inspection

Enabling In-Line Bottle Inspection with AI Edge Inferencing

With over 60+ years of operation, this in-line bottle inspection solution provider has been integrating their automated state-of-the-art bottle quality control vision systems to major beverage and food companies. They have successfully expanded and tapped into over 20 international territories with dedicated local support teams throughout.

The inspection solution is integrated onto high-speed production lines and ensures that each product has been examined and analyzed for defects before reaching the shelves of wholesale-retailers. The bottle inspection provider has developed their in-house interface that utilizes machine vision to inspect for empty bottle/cans, pressure closure, tamper bands integrity, high & low fill levels, and much more. By automating this inspection process, food and beverage companies significantly reduce the risk of delivering faulty products and maintain the brand reputation.

Challenge:

  • Sought out a reliable edge computing solution to streamline their machine vision system for in-line bottle inspection
  • Previous vendor made design layout changes without prior notification
  • Looking to upgrade from a near end-of-life Intel 7th Gen SBC solution to a modernized rugged industrial computer to qualify for EMC certifications
  • Consolidate previous system solutions into a singular computing platform to save on costs, efficiency, and space
  • Seeking a competent and communicative manufacturer local to California, USA for reliable engineering/sales support and supply chain management

    Solution:

    • Premio’s AI Edge Inference Computer, RCO-6000-CML Series, featuring high-performance processing and flexible connectivity engineered to power machine vision applications at the rugged edge
    • Configured with Intel XEON-W TE processor for server-grade performance with edge reliability and an extended product lifecycle roadmap
    • System consolidation by powering eight vision cameras simultaneously using EDGEBoost I/O modules
    • Fanless and cableless design that has undergone testing and validation to achieve world-class certifications (CE, FCC, UL) for safety standards compliance

      Benefits:

      • Premio is strategically located in California, USA for localized supply chain management
      • Dedicated engineering and sales team to address an on-going issues and future projects
      • Effective communication between Premio and solution provider with transparency and competency 

        The Challenge

         Bottle Inspection Challenges Icon List

        After several miscommunication and late notifications of a layout change for their open-board design, this bottle inspection provider sought out a new edge computing solution for their bottle inspection system. The company had been operating with an open-board SBC design that was enclosed in a cabinet until they discovered some fault with their vendor. The vendor constantly made unannounced design layout changes to their SBC and provided an inconsistent product that the bottle inspection provider constantly had to build around. This became a major issue and was the pushing factor in their search for a more reliable product from a competent and consistent partner.

        Another factor in their decision is that the SBC they were utilizing was powered by Intel’s 7th Gen processor which has become quite outdated and will soon to be EOL. The machine vision provider needed to find an edge computing solutions partner that will have modernized products and reliable long-term product lifecycle support.

        Although not actively seeking to consolidate, it would become a beneficial bonus if they could power 8 inspection cameras simultaneously through one computer, in comparison to their previous solution using two SBCs to power 4 cameras each. The machine vision provider was also in the phase of developing an EMC certified system which requires the edge computer to be fully ruggedized as compared to their previous SBC design.

        Lastly, the machine vision provider preferred to collaborate with a partner that is strategically located nearby to avoid supply chain disruptions and quick support turnaround if needed. Having a localized partner will greatly benefit their supply chain management to significantly reduce risks of supply chain disruptions, increase cost savings, increase communication and flexibility.

         

        The Solution

        Premio specified that the RCO-6000-CML Series, AI Edge Inference Computer, was best suited for their inspection vision system featuring seamless configurability and flexibility to meet their deployment specifications.

        RCO-6000-CML AI Edge Inference Computer for in-line bottle inspection vision AI application

        Leveraging 10th Gen Intel Processor for Machine Vision 

        The RCO-6000-CML Series is configured with a 10th Gen Intel XEON W1290TE processor to streamline the vision AI interface for bottle inspection while maintaining a 35W TDP. By configuring the machine vision provider with a TE model processor, it falls under Intel’s embedded product lifecycle roadmap for ensured long-term support and stability. Although vision AI workloads typically require a GPU for high-performance acceleration, the server-grade XEON processor is sufficient to drive their machine vision bottle application; increasing cost savings and maintaining a compact footprint.

        System Consolidation with EDGEBoost I/O Technology

        One of the most significant features of our AI Edge Inference Computers is the support for dual EDGEBoost I/O modules. EDGEBoost I/O allows our machine vision provider to configure their edge computing solution with a mix-and-match approach for near limitless I/O flexibility and compatibility.

        In this case, their bottle inspection system required support for 8 vision cameras. When compared to their previous solution, they utilized two SBCs to drive 4 cameras each. The RCO-6000-CML Series consolidates this by supporting the 8 cameras through EDGEBoost I/O modules all within a single unit. This system consolidation allows the machine vision provider to essentially save on space, software licensing, and overall costs.

        Fanless & Cableless Design for In-Line Edge AI Inferencing

        When deployed, the bottle inspection system will be integrated onto factory floors, operating in-line with the rest of production. The RCO-6000-CML Series is purpose-built with a fanless and cableless design for improved system reliability and durability. It has been tested and validated to hold up against the harsh conditions and environments of factory floors such as frequent shock and vibration, varying temperature changes, and unstable power fluctuations.

        As an additional benefit, the machine vision provider was in a design change phase to develop their system to qualify for EMC certifications. Premio’s RCO-6000-CML has undergone thorough testing and validation to achieve world-class certifications from CE, FCC, and UL for safety standards compliance. This, in return, had made the machine vision provider’s capabilities to achieve EMC a seamless effort.

         

        The Benefits

        In addition to our rugged edge computing solution, Premio has leveraged external factors that the bottle inspection provider found extremely beneficial. The partnership between the solution provider and Premio had proven such a success that they have referred Premio to their affiliate for consideration in future projects.

        Premio Inc State-of-the-art edge computing manufacturer located in City of Industry, California

        Robust Computing Product Portfolio

        Recognizing that there is no ‘one edge solution fits all’ approach, Premio has established a wide portfolio of off-the-shelf rugged edge computing solutions to meet all varieties of industrial deployment specifications. From ultra-compact industrial mini computers to machine vision computers that support dual FHFL GPUs, a reoccurring theme or feature within most of Premio's industrial computers is the ability to offer customizability with:

        • Modular EDGEBoost I/O: seamless plug-and-play connectivity and compatibility
        • EDGEBoost Nodes: tailored to meet specialized hardware acceleration  
        • Hot-swappable Storage: convenient data offloading procedures and upgradability
        • PCIe expansion: near limitless support for additional add-on cards

        Localized Supply Chain Management

        Premio is strategically located in City of Industry, California and has a dedicated supply chain team to expertly navigating through supply chain disruptions successfully. Being local to the machine vision provider leads to benefits such as decreased lead times and increased cost efficiency. As a result, Premio established a strong business relationship that enhanced the party's trust and delivery. 

        Engineering & Sales Support

        As solidified partners, Premio extends its support for any issues or concerns, as well as quick turnaround times for future projects or orders. The engineering and sales team has provided its expertise to ensure reliable operations and transparent communication to the machine vision provider.