Aurora AI by 極光有序有限公司

An order arrives find its drawings and verify revisions

Which drawings does an order need, and are any missing or the wrong revision? Aurora’s DWG and order-spreadsheet example explains prerequisites, number checks and delivery review.

Key takeaways

Define a traceable drawing-number key
Retain background-task state
Verify the output against both its source and merged result

This article uses an Aurora workflow demonstration

Aurora demonstrated placing DWG drawings and an order spreadsheet in one task, organizing drawing-number matches, previewing a merged PDF and downloading a ZIP. This illustrates one document workflow; it does not imply that all factories use this combination.

The underlying problem may be slow searches, missing files, unclear revisions or repeated attachment preparation. Data may instead live in a drawing-management system, ERP or other formats. Describe the current work before deciding whether this example applies.

The DWG/spreadsheet approach requires comparable drawing or part numbers, readable files, clear revision rules and an approved source. Without shared identifiers, establish mappings first. If an existing system handles the work reliably, another matching process may be unnecessary.

Matching examples: the same number is not enough for delivery

These four records are hypothetical teaching examples, not customer results. Compare the order number, required revision and candidate file separately, using a human-reviewed expected mapping for acceptance.

AB-001/R2 → AB-001-R2.dwg: number and revision match; verify the source before delivery. AB-002/R2 → AB-002-R1.dwg: revision mismatch; flag for review rather than substitute.

00123/R1 → 123-R1.dwg: do not remove leading zeros without an approved numbering rule. AB-004/R1 → no file: keep an unmatched record rather than silently omit it.

What a reviewable delivery manifest contains

Keep the order row, original drawing number, required revision, candidate file, matching method and review outcome. For normalization, retain both original and transformed values so mismatches can be traced.

Before export, confirm that matched, pending and unmatched records account for every input row. Then check PDF order and the ZIP manifest. Define how duplicate order rows are handled before using file counts as a success measure.

Prepare a sample set that an engineer can verify

Bring de-identified DWG files, an order spreadsheet, and an expected mapping reviewed by an engineer. For each row, retain the original drawing number, revision, expected file, and reason for any unmatched item. Include normal cases, missing files, duplicate numbers, and multiple revisions. This mapping is the acceptance baseline, not a system-generated answer.

Label drawing-number, quantity, and revision columns explicitly, and preserve leading zeros. Numeric spreadsheet cells may lose leading zeros before import. Resolve those issues at the source rather than guessing missing characters from similar identifiers.

Normalize formatting without erasing revision differences

This is a hypothetical rule example, not a customer case: if both parties confirm that letter case has no meaning, ab-001 in an order and AB-001.dwg can form a candidate match. AB-001-R1 and AB-001-R2 must not become the same approved revision simply because a suffix was removed.

Start with approved exact-matching and formatting rules. Missing matches, multiple candidates, and uncertain revisions require engineering review. Compare every result with the expected mapping, record incorrect or missed matches and manual corrections, then check the merged PDF order and ZIP contents.

Define the matching key before discussing automation

Automatic matching between a DWG and an order spreadsheet depends on a stable, recognizable drawing or part number. If the file name, title block, and order column use different formats, the workflow first needs normalization and exception rules.

Acceptance should therefore ask more than whether a match was found. It should show which source value produced each match, how unmatched items appear, and whether a reviewer can identify format or revision differences quickly.

Keep the inputs with the task

Engineering drawings and order data often come from different people and may arrive at different times. If only the final output remains, it becomes difficult to determine which drawing revision and which order file produced it.

Aurora's engineering-document example keeps the DWG, order spreadsheet, and processing task in the same working context. A reviewer can return to the inputs, match count, completion state, and deliverables instead of reconstructing context from a download folder.

A background task is more than a spinner

File parsing, conversion, and merging can take time. A useful background task exposes queued, processing, completed, or failed state and remains available after the person leaves the page.

Acceptance should deliberately test refresh, leave-and-return, incomplete input, and processing failure so task state does not disappear with the browser session.

Delivery is more than a download button

A verifiable delivery should answer three questions: which drawing numbers matched, whether the merged content is correct, and what the final download contains. Aurora currently demonstrates that path through match results, merged-PDF preview, and ZIP download.

If preparing order drawing files is repetitive and the prerequisites above hold, consider this example for a pilot. Verify reviewable outputs before deciding whether other systems need connecting.

Frequently asked questions

What should be checked first for DWG and spreadsheet matching?

Confirm the drawing- or part-number fields used on both sides, format differences, revision rules, and exceptions. Without a stable key, organize the data rules before deciding how much to automate.

What should engineering-document acceptance include?

At minimum: source files, matched and unmatched items, task state, a previewable merged result, and clearly identified downloadable output.

Does the first phase require ERP or MES integration?

Not necessarily. Aurora can first verify matching and delivery with de-identified DWG and order spreadsheets. Core-system integration should be evaluated separately after the workflow's success criteria are clear.

Connect the question to Aurora

The article provides a decision framework. Aurora's product, solutions, and workflow review put that framework into a real piece of work.