ABS.AI

RollOptix

A custom image-processing decision support tool that eliminates material waste and accelerates customer pricing for Egypt's leading F&B packaging manufacturers.

RollOptix is an AI-powered carton roll optimization engine built for industrial carton manufacturers. It transforms a slow, error-prone manual calculation into a real-time, validated decision automatically selecting the optimal roll width for every new carton design to minimize material waste and accelerate customer pricing offers.

RollOptix was built for a leading Egyptian food and beverage packaging company to automate carton material optimization. Instead of manually calculating the best raw material roll width for each carton design, the system instantly analyzes available widths and returns the lowest-waste option with a full comparison across all configurations accelerating pricing and quotation workflows.

The main challenges included achieving millimeter-level calculation accuracy, optimizing both horizontal and vertical carton orientations, handling trim and rotation edge cases, designing an interface usable by both junior estimators and production managers, and integrating the optimization results directly into the pricing workflow to speed up customer response times.

  • Design Dimension Input Module: Single-screen entry of the carton design's width and height with optional trim and bleed parameters.
  • Roll Library Module: Manages the three default market widths (35 / 70 / 140 cm) and custom widths with per-meter pricing.
  • Image Processing Layout Engine: The core algorithm calculating how the carton design tiles against each candidate roll width and quantifying leftover material.
  • Loss Comparison Module: Side-by-side breakdown of material loss for every default roll width shown as both linear waste and percentage.
  • Layout Visualization Module: Renders the carton layout on the chosen roll for visual confirmation of how the design will be cut.
  • Orientation Solver Module: Automatically evaluates horizontal and vertical placements, selecting whichever orientation yields the lowest material loss.
  • Optimal Width Recommendation Module: Surfaces the lowest-loss option as the recommended choice with a plain-language explanation.
  • Manual Override Module: Allows entry of any custom roll width with instant recalculation against the defaults.
  • Pricing Handoff Module: Exports the chosen roll width, loss percentage, and material cost into the pricing offer workflow with one click.
  • Calculation History Module: Stores every calculation against the customer and design, building a reusable archive for repeat orders and audit trail.

RollOptix delivers a focused set of features designed to compress the carton roll selection process from hours to seconds while preserving full human oversight.

  • Single-input dimension entry for any new carton design.
  • Automatic comparison across all three default roll widths (35 / 70 / 140 cm).
  • Horizontal vs. vertical orientation auto-evaluated for every candidate width.
  • Real-time material loss percentage per roll width in linear and percentage view.
  • Manual custom-width override with instant recalculation against defaults.
  • Visual preview of the carton tiling pattern on the selected roll.
  • One-click handoff to the pricing offer workflow.
  • Calculation history for repeat orders and full audit trail.
  • Mobile-friendly interface for use on the production floor.
  • Real-time alerts notifying engineers when stock reaches critical levels.

Platform Architecture

RollOptix is a purpose-built AI Decision Support Tool designed to remove a critical bottleneck in the client's pricing-to-production cycle. The platform operates across three interconnected layers:

  • Input Layer:: A clean, single-form interface where the estimator enters the carton design's dimensions and any trim or orientation constraints. Built for speed — most calculations begin with under ten seconds of data entry.
  • Computation Layer:: An image-processing engine that solves the carton-on-roll tiling problem across every candidate width and orientation in parallel, quantifying material loss to the millimeter for each option.
  • Decision Layer:: A validation dashboard where the production manager reviews the recommendation, compares all default widths against custom alternatives, optionally enters a manual override, and pushes the final result directly into the pricing offer workflow.

All three layers are unified under a single deployment, ensuring every calculation is traceable, every override is logged, and every pricing offer is backed by a defensible optimization run.

  • 65% higher material estimation accuracy.
  • 60% lower cost overruns.
  • 55% improved operational efficiency.
  • Next
  • Node js
  • Python
  • Express
  • Mongo db
  • contabo