Skip to content
Eli Lewis
Eli Lewis · elilewis2029@u.northwestern.edu · elilewisportfolio.info/work/detail-room-storage-plan-and-barcode-prototype

Engineering · 2026

Detail room: $0 storage plan and barcode lookup prototype

Room layout in SketchUp and a working inventory mockup (React, SQLite, Code 128 labels)

Searching "6203 bearing" finds the part, its total stock and both bins it's linked to (sample data).

Searching "6203 bearing" finds the part, its total stock and both bins it's linked to (sample data).

I planned a small detail room for dense bin storage using only equipment the shop already owned, and designed a barcode workflow so any employee could find a part. Management approved the layout and agreed with the barcode approach, room construction kept it from being installed before I left. Part finding is easy: fuzzy search by SKU, size, type, description, with filters available. Part locations are indexed by Unit-->Row-->Column-->Bin color for quick retrieval of parts. System was designed to be intuitive and naturally error-resistant: every action starts with scanning a part, every time. Every action requires a deliberate button press on the scanner, accidental scans will not change data. The user is always in charge, the system will not make forcing decisions: Parts can be linked to multiple bins or multiple parts to a bin, just with a warning; overdraw is allowed in case of previous count error, unlimited labels can be made with no location links at all

Role
SoloI was responsible for the room plan, the workflow design and refining the mockup, the code was made with AI
Duration
Summer 2026
Tools
chatgpt-codex
Skills
inventory-systems, workflow-design, facility-layout, prototyping, human-centered-design

Goal & constraints

  • Fit dense storage into a small room, with every part findable by anyone.
  • High-variety, low-quantity parts: depleted items often aren't restocked, so bins get reused.
  • Use storage solutions the shop already owns
  • Ambiguity-proof: A system that leaves someone unsure of the next step wastes time invites errors, the design must preemptively predict those scenarios and create a clear course of action, such as a dedicated location for intake parts someone might not have had time to sort then and there

Process

-- Designed the workflow: scan a part, Take is the default, quantity starts at 1, confirm. No employee badge or job number.

  • Made location changes deliberate: each bin has a permanent location label, and a part is linked to a bin only when someone scans both.

  • Created a working mockup with AI: search a part to see its total and bins, link by scanning, a front view of each rack, working barcode creation and printing, and a history log. Scanner and printer are simulated on screen.

  • Showed it to the lead engineer's assistant, who currently manages inventory (highly positive feedback)

  • Planned for budget MVP: DIY handheld made with a WIO terminal, button hardware, and a scanner component

  • Planned the room in SketchUp: two rolling Uline bin racks, three LISTA cabinets and a computer table, all equipment the shop already owned. A third rack would add $760, or $900–1,090 with bins.

  • Refined a working mockup built with ChatGPT Codex: (search, scan-to-link, rack map, label printing, history log; scanner and printer simulated) and priced budget hardware, from USB scanners and old Android phones to a DIY Wio Terminal + Zebra LS1203 handheld.

Before: the small computer room, old office and hallway
Before: the small computer room, old office and hallway
The planned detail room (left), using only equipment the shop already owned: two rolling bin racks that wheel out for full access, three LISTA cabinets, a computer table, and a spot for a cart or filing cabinet.
The planned detail room (left), using only equipment the shop already owned: two rolling bin racks that wheel out for full access, three LISTA cabinets, a computer table, and a spot for a cart or filing cabinet.
Scan and place: scan the part, then the bin's location label, then confirm. The scans are simulated on screen but go through the same lookup a real scanner would use.
Scan and place: scan the part, then the bin's location label, then confirm. The scans are simulated on screen but go through the same lookup a real scanner would use.
Storage map: a front view of each rack; click a bin to see which parts are linked to it. New bins can be added to open spots
Storage map: a front view of each rack; click a bin to see which parts are linked to it. New bins can be added to open spots

Result

$0 base layout using owned equipment and finalized scanning workflow

What I’d change

I knew this was the right idea, yet I waited too long to get approval from my managers to work on a mockup. If done again, I would have created this weeks earlier, giving me time to setup a working MVP.

Printing this page gives a one-to-two page PDF.