Hugdrigo's picture
Upload 3 files
11d043a verified
|
Raw History Blame Contribute Delete
5.73 kB
metadata
title: BOMForge | Manufacturing Requirements Studio
emoji: 馃彮
colorFrom: blue
colorTo: green
sdk: static
app_file: index.html
pinned: false
license: mit
short_description: Production planning and materials simulation.

BOMForge 路 Manufacturing Requirements Studio

Hugdrigo | Project #113

BOMForge is an interactive, browser-only material requirements planning (MRP) simulator. It helps operations teams explore how production demand turns into component requirements, planned purchases, stock projections, and release-timing warnings. The initial example is synthetic and educational.

Run and publish

  1. Create a new public Space at https://huggingface.co/new-space with owner Hugdrigo and repository name bomforge-manufacturing-planner.
  2. Select Static (blank template). Choose MIT if prompted.
  3. Extract the ZIP and upload all three files directly into the repository root: README.md, index.html, and LICENSE.
  4. Commit the files. The Space reads the YAML at the beginning of this README and serves index.html. No Python, Gradio, credentials, GPU, build step, or npm dependencies are required.
  5. Confirm the app loads, open the MRP planner, change the scenario, then inspect the materials and downloads.

All thumbnail colors are supported values, colorFrom: blue and colorTo: green, as documented in the official Spaces configuration reference.

Highlights

  • Overview: production requirements, 8-week outlook, raw-material purchase estimate, early-release queue, and cost concentration.
  • MRP planner: configurable demand and safety stock, baseline / surge / delay / buffer scenarios, time-phased gross requirements, projected inventory, planned receipts and planned releases.
  • Materials: edit existing SKUs or create new ones; register one parent BOM link for a new material, choose quantity, update lead time, initial stock and costs. Archive and restore without deleting existing records.
  • Scenario lab: compare four scenarios against baseline, modify weekly finished-product demand, evaluate purchase costs and early-release warnings.
  • Methodology: the transparent calculation rules, assumptions, and important limitations.
  • Exports: material summary CSV, detailed time-phased MRP CSV, and complete JSON backup. Import accepts versioned JSON backups and rejects invalid structures.
  • Design: responsive navigation, light and dark themes, keyboard-friendly controls, built-in SVG chart, and no external libraries.

Model structure and calculations

The demonstration includes 2 finished goods, 2 subassemblies, 6 raw materials, and 13 parent-component relationships. A topological sort processes parents before children, rejecting cyclic active BOM configurations. For each SKU and each week:

gross_requirement = independent_finished_product_demand
                 + demand_exploded_from_parent_releases
planned_receipt   = max(0, gross_requirement + safety_stock_target - available_inventory)
ending_inventory  = available_inventory + planned_receipt - gross_requirement
release_week      = receipt_week - lead_time

For gross demand coming from a parent, child demand is scheduled when the parent order is released. If a parent release date falls before week 1, its release is flagged as pre-horizon and its component requirements are assigned to the first modelled week. Pre-horizon flags do not guarantee an order can be delivered on time.

The raw-material cost estimate is the sum of planned raw-material receipts multiplied by synthetic unit prices. Finished-goods forecasts are rounded to whole units. Scenario assumptions: surge multiplies final demand by 1.35, delay adds one week to raw lead times, buffer multiplies the safety target by 1.5. User-selected demand and safety adjustments apply multiplicatively.

Data handling and limitations

  • All inputs and names are synthetic, not actual business records, quotes or forecasts.
  • Results are planning illustrations, not supplier commitments or guaranteed delivery dates.
  • No modeling of manufacturing yield, labor, machine capacity, calendared shifts, supplier minimum-order quantities, volume discounts, taxes, freight, open orders, backorders or stochastic lead times.
  • Inventory receipts are assumed on time and sufficient to cover calculated requirements; inventory is netted deterministically.
  • Browser storage keeps edits only on the current browser/device. Use a JSON backup to transfer a workspace; backups are not encrypted.
  • This standalone static app makes no network requests and has no account creation, telemetry or API dependency.
  • Imported data is constrained by finite numeric bounds, maximum item counts, valid SKU identifiers and acyclic active BOM relationships. Text is escaped in HTML and dangerous spreadsheet formula prefixes are escaped during CSV export.

Quality assurance

  • Tested with automated Chromium interactions on the major screens, including editing, archiving, restoring, creating materials, cycle rejection, scenario changes, exported CSV/JSON, JSON import, malformed-data rejection, local persistence and theme changes.
  • Compared the JavaScript planning model against an independent Python implementation over 1,200 randomized scenario cases with 99,600 field-level numeric checks.
  • Responsive checks included 320, 360, 375, 390, 414, 480, 600, 768, 1024 and 1440-pixel viewports.
  • Metadata was parsed and checked against the official static Space rules. These tests do not replace verification after deployment.

Copyright and licensing

MIT License. Copyright (c) 2026 Hugdrigo. See LICENSE.