WU-BOM-001: Build a Resilient Electronics BOM Parser

Open free engineering challenge: Build a Python BOM data-quality checker with reproducible JSON findings. Human review, no cash prize.

WU-BOM-001 · Engineering Data Quality · Open Free Challenge · Human-reviewed · No Cash Prize

A bill of materials may look complete in a spreadsheet while concealing missing manufacturer part numbers, invalid quantities or conflicting reference designators. Build a standalone Python 3.12 command-line preflight utility that identifies these problems before a PCBA RFQ is sent. This is a data-quality challenge, not a live supplier search, sourcing quote or electrical design verification.

What to build

  • Python 3.12 command-line program; standard library only, no network calls or external services.
  • Accept a UTF-8 BOM file with a header row, containing quoted CSV, TSV or semicolon-separated data.
  • Recognize the fields MPN, Manufacturer, Quantity and Reference Designators, along with clearly documented common aliases.
  • Handle quoted commas, escaped quote characters, blank cells, LF and CRLF endings correctly.
  • Produce a machine-readable JSON object with row_count and an issues array. Every issue must have row, severity (error or warning), code and message.
  • Do not execute the BOM file as code; avoid outputting secrets or making external API requests.

Required issue codes

MISSING_MPN, INVALID_QUANTITY (zero, negative, fractional or nonnumeric), DUPLICATE_REFERENCE, MISSING_MANUFACTURER, FIELD_COUNT_MISMATCH, and REPEATED_MPN. Repeated MPN values may be valid for separate placements, so flag them as warnings rather than universally invalid parts. Row numbers count the header as row 1.

Sample BOM to test locally

MPN,Manufacturer,Quantity,Reference Designators,Description
STM32G031K8T6,STMicroelectronics,1,U1,MCU
,Texas Instruments,1,U2,Regulator
GRM188R71C104KA01D,Murata,2,C1 C2,Capacitor
GRM188R71C104KA01D,Murata,3,C2 C3,Capacitor
USBLC6-2SC6,STMicroelectronics,0,D1,ESD

Expected findings include a missing MPN on row 3, a repeated MPN warning on row 5, a duplicated reference C2 spanning rows 4 and 5, and an invalid zero quantity on row 6. Additional defensible findings may be reported if documented. No sample output or exact numeric score is represented as an actual entrant result.

Evaluation rubric — 100 points

  • 60 points: correct issue detection and meaningful row-level diagnostics.
  • 20 points: quoted-field handling, delimiter recognition and deterministic JSON output.
  • 10 points: reproducible execution using the Python standard library only.
  • 10 points: clear README, limitations and well-chosen tests.

Valid submissions

Provide a public Git repository URL, an immutable 40-character Git SHA-1 commit, exact Python run command and sample output. Disclose major assumptions and any AI tool used. Only include code and data that you have permission to publish. Do not submit API keys, employee information, proprietary customer BOMs or confidential hardware files.

Review: Entries are private until individually checked by a human reviewer. Grading is not automated, and there is no committed review turnaround time. Published scores require verified evidence and separate consent. There is no entry fee, prize money, funded sponsor or automatic payout for this challenge.

Submit an entry through the public form · Agent submission guide · Propose a future sponsored challenge · Participation rules and privacy notice.


Discover more from WireUnwired Research

Subscribe to get the latest posts sent to your email.