Detailed Explanation of Fang Xiaohui's Three Major AI Functions: AI Quotation + AI Process Card + AI

2026-06-23

In the wave of digital transformation in manufacturing, discrete manufacturing and hybrid manufacturing enterprises are facing unprecedented challenges——surge in demand for multi-variety, small-batch, and customized products, complex and ever-changing process routes, quoting relying on the experience of “master craftsmen”, and scheduling resembling a “guessing game”. As a national-level “specialized and new” “little giant” enterprise deeply rooted in the manufacturing informatization field for over 20 years, Fangtian Software proudly launches the AI intelligent robot “Fang Xiaohui”, which, with its three core functions—AI Quotation, AI Process Card, and AI Intelligent Scheduling—builds an intelligent decision-making closed loop from sales to delivery for discrete and hybrid manufacturing enterprises.


I. Three Major Pain Points of Discrete and Hybrid Manufacturing

Discrete manufacturing (e.g., molds, hardware machinery, electronic assembly) and hybrid manufacturing (combining discrete and process characteristics, e.g., plastic molding) enterprises commonly face three dilemmas:

  • Difficult Quoting: Non-standard customized products require manual disassembly of drawings, estimation of materials, prediction of process routes, and calculation of working hours. The quoting cycle can take days, and it is highly prone to the awkward situation of “quoting too high and losing the order, or quoting too low and incurring losses.”

  • Slow Process Card Preparation: Process cards are the “legal documents” guiding production, but traditional preparation heavily relies on the personal experience of “master craftsmen.” It often takes 3 days to prepare a single process card, and staff turnover leads to the loss of process knowledge.

  • Scheduling Relies on Experience: Facing multi-variety, small-batch production scenarios, traditional Excel manual scheduling or simple-rule ERP planning modules suffer from low resource utilization, slow response to order insertions, and frequent delivery delays.

“Fang Xiaohui” was born precisely to solve these pain points.


II. AI Quotation: Second-Level Response, Letting Business Opportunities Not Wait

Under the traditional quoting model, engineers face complex drawings and need to manually disassemble features, estimate materials, predict process routes, and calculate working hours, often taking days to provide a quote. “Fang Xiaohui’s” AI Quotation function completely changes this situation:

  • Intelligent Drawing Analysis: Combining computer vision with large language models, it directly reads CAD/3D drawings, automatically identifying geometric features of parts (e.g., holes, slots, curved surfaces), material requirements, and precision levels.

  • Historical Case Matching: Based on the massive historical order database accumulated by the Fangtian T-ONE system, it instantly retrieves processing cases of similar parts, referencing their actual material consumption, working hours, and equipment types.

  • Dynamic Cost Modeling: It incorporates a real-time updated raw material price database, equipment rates, and labor cost models, automatically generating detailed quotations including material costs, processing fees, management fees, and profit.

Results: The quoting cycle is compressed from “days” to “minutes,” and quotation accuracy improves by over 30%. According to Fangtian’s internal test data, after introducing AI automatic quoting, the order loss rate due to estimation errors decreased by 90%.


III. AI Process Card: Digitizing and Standardizing Master Craftsmen’s Experience

The preparation of process cards (SOP) is one of the most time-consuming and experience-dependent links in discrete manufacturing. “Fang Xiaohui’s” AI Process Card function achieves a leap from “manual preparation” to “intelligent generation”:

  • Zero-Baseline Automatic Generation: Simply input basic information such as the workpiece name, specifications, and material, and it can generate a complete process route with one click, covering core content like processing steps, equipment selection, and cutting parameters.

  • Driven by Process Knowledge Graph: It transforms industry-standard processing rules and the enterprise’s internal tacit knowledge (e.g., heat treatment deformation patterns for specific materials, optimal cutting parameters for specific equipment) into a structured knowledge graph.

  • Explainable Recommendations: It not only provides the process plan but also explains “why this step is chosen” and “why this equipment is used,” and offers alternative plans for engineers to review and confirm.

  • Continuous Self-Evolution: As production data flows back (e.g., actual processing time, yield feedback), it continuously revises and optimizes the process recommendation logic.

Results: Process preparation efficiency improves by 80%, reducing the work that originally took 3 days to within 2 hours. Actual data from a die-casting mold factory shows that after the upgrade, process design labor costs decreased by 72%, and the order delivery cycle was compressed by 34%.


IV. AI Intelligent Scheduling: From “Experience Estimation” to “Data-Driven Calculation”

Production scheduling is the most troublesome “black box” problem for manufacturing enterprises. “Fang Xiaohui’s” AI Intelligent Scheduling function brings a revolutionary breakthrough:

  • Intelligent Reverse Scheduling: It pioneers an automatic back-calculation mechanism based on the predetermined completion date and the proportional allocation of process time. Without needing to pre-enter standard working hours, the system can accurately decompose the standard working hours of each process card and automatically derive the planned start and completion times.

  • Real-Time Capacity Constraint Optimization: Combining factory equipment status (e.g., number of idle machines) and tool life data, it dynamically adjusts the sequence and parameters of processes to maximize resource utilization.

  • Multi-Scenario Simulation Scheduling: Based on real-time capacity, material constraints, and order priorities, it performs multi-scenario simulation scheduling to generate executable and optimizable production plans.

Results: Scheduling moves from “shooting from the hip” to “data-driven,” significantly improving resource utilization and greatly reducing delivery delays.


V. Synergy of Three Functions, Connecting the Entire Chain from Quotation to Delivery

The three AI functions of “Fang Xiaohui” do not exist in isolation but form an organically synergistic intelligent closed loop:

  • AI Quotation: Accurately locks in the cost and profit margin of orders, providing data support for order acceptance decisions;

  • AI Process Card: Transforms the process predictions from the quotation stage into executable standardized work instructions, ensuring “accurate quoting and feasible execution”;

  • AI Intelligent Scheduling: Based on the working hour data from process cards and real-time capacity, it generates the optimal production plan, ensuring “smooth scheduling and timely delivery.”

These three elements are interlinked, helping enterprises achieve a leap in efficiency across the entire chain, from rapid quoting at the sales end, to process planning at the technical end, and to intelligent scheduling at the production end.


Under the dual dilemma of “data silos everywhere” and “gaps in expert experience,” discrete and hybrid manufacturing enterprises need not just a set of management software, but an intelligent decision-making center that “understands the business, can think, and can execute.” Fangtian Software’s AI intelligent robot “Fang Xiaohui,” with its three core functions—AI Quotation, AI Process Card, and AI Intelligent Scheduling—is helping more and more manufacturing enterprises move from “sweat-driven” to “wisdom-driven,” making manufacturing smarter and decisions more precise.

Smart Manufacturing with a Method, Boundless World——Fangtian Software, Creating a New Future of Intelligent Manufacturing with You.




share
Next:This is the last one
Prev:This is the first article