Add research workflow guide

README.md

@@ -1,3 +1,90 @@
-# patent-standards-research-trail
+# Patent and Standards Research: Keeping a Trail
 
-Patent and Standards Research: Keeping a Trail
+Patent and standards documents are some of the most demanding sources a researcher can work with. They are long, written in specialised language, cross-referenced heavily and legally significant. Engineers use them to understand the state of the art, product teams use them to check freedom to operate, standards experts use them to track how specifications evolve, and analysts use them to map competitors' technical strategies. In all these uses, the conclusions often carry real consequences: a design decision, a licensing negotiation or an investment. That makes keeping a clear trail from every conclusion to its supporting passage essential. This guide explains how to research patents and standards in a way that remains traceable and reviewable.
+
+## Why trails matter more here
+
+Three features of patent and standards research make traceability especially important:
+
+- **Precision of language.** Small differences in wording, such as "comprising" versus "consisting of" in a patent claim, or "shall" versus "should" in a standard, change meaning substantially.
+- **Versioning.** Patent applications are amended during prosecution, and standards go through drafts, revisions and corrigenda. A statement true of one version may not be true of another.
+- **Legal and commercial stakes.** Conclusions may be scrutinised by lawyers, partners or regulators who need to see exactly what was relied on.
+
+A trail that records which document, which version and which passage supports each conclusion protects both the quality of the analysis and the people who rely on it.
+
+## Planning the search
+
+Begin, as with any research, with a clear question. Common question types include:
+
+- **State of the art**: what approaches to a technical problem already exist?
+- **Patentability**: is a proposed invention novel and non-obvious relative to existing disclosures?
+- **Freedom to operate**: could a product infringe active patents in particular jurisdictions?
+- **Landscape**: who is patenting in a field, and how is activity changing?
+- **Standards compliance**: what does a given standard require for a particular feature?
+
+Each type implies different sources, date ranges and levels of rigour. Freedom-to-operate questions, for instance, typically require professional legal input; the research trail you build supports that input rather than replacing it.
+
+## Search techniques
+
+Patent searching combines keywords with classification codes. Classification systems group patents by technical subject regardless of the words used, which matters because patent drafters often use unusual terminology. A good strategy uses both: keywords to find initial relevant documents, then the classification codes of those documents to find others, then citation searching backward and forward from the most relevant patents.
+
+For standards, start from the official catalogues of the relevant standards bodies, check the current status of each document, and identify related documents such as amendments, test methods and normative references that the main standard depends on.
+
+Record every search: database, date, query, classification codes, filters and number of results. This record is the first part of your trail.
+
+## Reading patents efficiently
+
+Patents have a predictable structure: bibliographic data, abstract, description, drawings and claims. For most research purposes:
+
+- read the **claims** to understand what is legally protected, starting with the independent claims;
+- use the **description** and **drawings** to understand how the invention works and how claim terms should be interpreted;
+- check the **legal status** and **family members** to understand where and whether protection is in force.
+
+When noting a finding, record the publication number, the specific claim or paragraph number and the version, for example the granted patent rather than the application.
+
+## Reading standards efficiently
+
+Standards distinguish between normative content, which defines requirements, and informative content, which provides explanation or examples. Note which kind of clause supports each finding. Record the standard number, edition, year, clause number and any amendments that affect the clause. Where a standard refers normatively to another document, that document becomes part of your evidence base too.
+
+## Building the trail
+
+For each conclusion, keep a record that includes:
+
+- the conclusion in one sentence;
+- the supporting documents with exact identifiers and versions;
+- the specific claims, paragraphs or clauses relied on, quoted where wording matters;
+- your interpretation and any assumptions;
+- the date of the research and the legal status at that date.
+
+Store copies of the documents you relied on, in the versions you used. Online documents can change status, and having the exact version available avoids later confusion.
+
+Research workbenches designed for evidence traceability make this much easier to maintain. Yila (https://yila.ai) keeps notes, summaries and comparison tables linked to the passages they were built from, so a conclusion such as "the prior art discloses a similar sealing mechanism" can be traced in one step to the paragraph and figure that disclose it.
+
+## Comparing documents
+
+Claim charts and feature matrices are standard tools for comparing a product or invention with patents or standards. Each row typically represents a claim element or requirement, and each column a document or product. Every cell should cite the specific passage that supports it. A claim chart without passage-level citations is an opinion; with them, it is evidence that others can check.
+
+## Keeping research current
+
+Patent and standards landscapes change continuously. New applications publish weekly, patents lapse or are invalidated, and standards are revised. For ongoing projects:
+
+- set alerts on key classification codes, assignees and standards;
+- record the date of each status check;
+- re-check legal status before relying on conclusions for decisions;
+- version your analysis, noting what changed and why.
+
+## Working with non-English documents
+
+Many relevant patents and standards are published in languages other than English. Machine translation is now good enough for screening, but key passages relied on for conclusions deserve a careful human check, because technical terms and claim language are easily mistranslated. Record the original-language passage alongside the translation in your notes, and note who checked it.
+
+## Common pitfalls
+
+- **Relying on abstracts.** Abstracts are not legally operative and often omit key limitations.
+- **Ignoring versions.** Citing an application when the granted claims differ, or an old edition of a standard.
+- **Keyword-only searching.** Missing relevant documents that use different terminology.
+- **Unrecorded interpretation.** Treating your reading of an ambiguous claim term as fact without noting the assumption.
+- **Untraceable charts.** Claim charts that do not cite specific passages.
+
+## Conclusion
+
+Patent and standards research rewards precision and punishes vagueness. By planning searches carefully, combining keywords with classifications and citations, reading documents with attention to their structure, and recording versions, passages and interpretations for every conclusion, you create a trail that colleagues, lawyers and partners can follow and trust. In a field where a single word can matter, that trail is the most valuable output of the research.