Food and Beverage
AI enhances traditional quality inspection to reduce errors, ensure food safety and avoid product recalls, and automate manual tasks. With new sensor technologies, such as hyperspectral imaging, manufacturers can add more advanced inspection for requirements including grading and product characterization, verifying product mix, and screening for foreign materials. Pleora’s AI solutions have been tested and deployed in inspection systems for labelling and packaging, grading and sorting, verifying fill levels and ingredient mix, and quality assurance.
With Pleora’s eBUS AI Studio platform, non-experts can design AI and computer vision plug-in algorithms without needing specialized development skills. Similarly, developers can use the platform to rapidly prototype and test plug-ins, mix computer vision and AI skills, customize open source models, and streamline production deployment.
AI plug-ins developed in eBUS AI Studio are optimized for Pleora’s AI Gateway edge processing platform. In a production environment, the edge processor intercepts the video and “mimics” the camera, which lets manufacturers evolve towards advanced AI inspection without disrupting installed infrastructure and processes. AI processed data is then sent over GigE Vision to the inspection application, which receives it as if it were still connected directly to the camera. Data can also be used for continuous offline plug-in training.
With a hybrid approach that adds AI capabilities alongside existing, proven machine vision processes, consumer product manufacturers can speed inspection by reducing false-positives that required costly and time-consuming manual secondary inspection. Inspection data can also be used for more comprehensive Industry 4.0 initiatives, with full-factory analytics to help improve processes and support proactive machine maintenance to reduce downtime.
For products requiring further manual inspection, Pleora’s visual inspection system guides operators through manufacturing steps and automatically highlights deviations and quality issues.
Adding Hyperspectral Inspection
Pleora’s off-the-shelf hyperspectral plug-in simplifies training and deployment for more advanced inspection. Where traditional cameras capture images in wavelengths to match human vision, hyperspectral imaging provides ultraviolet or near infrared information for each pixel on an inspected object.
With the hyperspectral plug-in, designers and end-users can deploy machine learning hyperspectral capabilities without any additional programming knowledge. Images and data are uploaded to ‘no code’ training software on a host PC, which automatically generates an AI plug-in that is deployed on the Pleora AI Gateway.
Pleora’s AI Gateway works seamlessly with any standards-compliant hyperspectral sensor, meaning end-users can avoid vendor lock-in while maintaining processes and analysis software. The AI Gateway automatically handles image acquisition from the hyperspectral imaging source and sending out the processed data over GigE Vision to inspection and analysis platforms.