Sorting and feeding line InVision²

Overview
InVision² is a fruit sorting and grading platform equipped with LUCAi™ deep learning software for high-precision optical grading. Designed for packhouses and processors, the system supports consistent grading across batches, varieties and operators while enabling improved pack-out and reduced fruit loss. LUCAi™ provides plug-and-play pretrained models and tools to adapt grading severity to seasonal and market requirements.

Benefits

• Precise grading: deep learning models and a pretrained model library developed with growers and packers to improve classification accuracy.
• Operational consistency: automated results reduce dependence on manual quality control and operator experience.
• Yield and pack-out improvement: optimized sorting reduces waste and increases usable yield for packing lines.
• Ease of integration: user-friendly LUCAi™ interface lowers training needs and accelerates deployment.

Cherry grading capabilities
The InVision² with LUCAi™ package includes ready-to-deploy models tuned for cherries and multiple varieties. Key capabilities include:

• Detection of edge cracks, pacman-shaped defects and open sutures.
• Identification of cosmetic blemishes, stem pulls and improved spur detection.
• Adjustable grading severity to reflect seasonal dynamics and market preferences without requiring highly specialised operators.
• Maintains consistent grading across seasons, lots and operators to prevent accidental downgrades.

Associated technologies

• LUCAi™ deep learning models (built into new InVision² platforms; available as an upgrade for existing InVision² cherry graders).
• Advanced camera-based imaging with pulsed LED illumination for color, size, shape and defect detection.
• Integration compatibility with other TOMRA food platforms and peripherals (examples: Small Fruit Sorter (SFS) and CURO filling systems).

Use cases and scope
Intended for high-precision optical sorting and grading of small fruits (demonstrated for cherries and referenced for apples, blueberries and citrus), InVision² with LUCAi™ targets packhouses and processors aiming to maximise yield, improve pack-out uniformity and minimise fruit loss while maintaining varietal and seasonal consistency.

Technical features / specifications

• Platform: InVision² grading platform with integrated LUCAi™ deep learning software.
• AI: Pretrained model library enabling plug-and-play deployment and continuous model updates based on grower/packer data.
• Defect detection: edge cracks, pacman cherries, open sutures, cosmetic blemishes, stem pulls, improved spur detection.
• Grading control: adjustable severity parameters to match seasonal and market requirements.
• Operation: intuitive LUCAi™ interface for reduced operator training and faster line integration.
• Performance goals: increased yield, optimised pack-out, minimal fruit loss and consistent grading across batches and operators.
• Upgrade path: LUCAi™ available as an upgrade for existing InVision² cherry graders.
• Imaging: camera systems with pulsed LED illumination for colour/shape/size/defect analysis.
• Compatibility: references include Small Fruit Sorter (SFS) and CURO systems; scalable to various infeed and peripheral configurations.
• Media: product video and brochure (PDF) available for technical details.