ERTH Router Retest & Distribution
ERTH Router Retest & Cleaning Operations Management System
Operational Workflow & State Machine
End-to-End Pipeline- 01Supervisor logs in via Google OIDC and selects staff member to assign a batch of incoming routers
- 02System increments cleaner's open allocation count and records timestamped distribution event
- 03Cleaner completes physical cleaning, adhesive removal, and sanitization at their workstation
- 04Cleaner returns batch to Return & Sort station; supervisor or lead scans serials via USB barcode scanner or camera
- 05System matches router model to point level (L1=2 to L4=10), credits points to cleaner's daily tally, and updates production scoreboard
RPOMS ecosystem
Separate systems- Flagship system, same programmeRPOMS
The router refurbishment operations platform: acceptance, packing, delivery and daily reporting. This app runs beside it on its own database and never connects to it.
Read its case study
The Operational Challenge
In the router retest and cleaning facility, routers arrive in vastly different physical conditions: some require only dust removal and wiping (Level 1: 2 points), while others arrive coated in stubborn adhesive, damaged stickers, or grime requiring solvent baths and scraping (Level 4: 10 points).
Without a tiered points system, staff were incentivized to cherry-pick easy routers and avoid difficult units. Tracking this on paper work logs led to disputes, lost tallies, and delayed end-of-shift reporting.
Architecture & Independence
Rather than overloading RPOMS with fine-grained cleaning sub-tasks, ERTH Router Distribution was architected as a fully independent Next.js 15 application. It operates with its own dedicated PostgreSQL database and Google OIDC PKCE authentication.
This architectural isolation ensures that high-volume floor workstation polling and real-time scanning do not impact RPOMS core inventory locks or Supabase connection pools.
Workstation Ergonomics: Dual-Mode Scanning
The Return & Sort station was engineered for maximum technician throughput. It supports both standard USB keyboard-wedge 2D barcode scanners and browser-based webcam optical scanning via the HTML5 Barcode Detection API.
As units are scanned, audio cues confirm successful registration and update the cleaner's live points progress bar against their 350-point target without requiring mouse or keyboard interaction.
What's live, and what isn't
“It exists in the repository” and “it runs in production” are different claims. This is the difference, stated rather than left to be assumed.
- ImplementedSupervisor distribution moduleFast bulk assignment without serial bottlenecking.
- ImplementedReturn & Sort inspection stationDual hardware scanner and camera barcode detection with instant point attribution.
- ImplementedWeighted productivity point engine (L1–L4)Transactional point ledger tracking daily benchmarks against 350-point quota.
- ImplementedGoogle OIDC PKCE authenticationRole-based authentication isolating supervisor actions from cleaner views.
- ImplementedProduction deployment handoffProduction container configuration and dedicated PostgreSQL migration prepared.
Technical stack
Frontend & UI
Backend & Database
Auth & Security
Limitations
- Requires stable network connection to the local floor PostgreSQL server.
- Camera-based barcode scanning requires adequate workstation lighting; USB hardware scanners are recommended for high volume.
- 6 modified files currently uncommitted in local working tree.
What's next
- 01Automated daily summary report export to Excel/PDF
- 02Integration with floor overhead display monitor for live productivity rankings
Source and further reading
Built for ERTH operations. Operational metrics and architectural patterns are documented; internal staff names and customer identification are protected.
Finished the technical breakdown?
Return to the concise HR overview or discover other projects in the portfolio.