How to Write an Invention Disclosure (Step-by-Step)

Not legal advice. Review with a registered patent attorney.

Step 1 — Define the Problem and Prior Art

• State the problem precisely (metrics, costs, failure modes). Provide baselines.
• List closest approaches (papers, patents, products) and their limitations.
• Tip: include links or IDs so counsel can quickly verify.

Step 2 — Describe the Core Idea and Advantages

• What is novel? Be specific about components/logic/materials that make the difference.
• Quantify advantages (latency, accuracy, durability, safety, cost). Provide ranges.
• Contrast with prior art to highlight the inventive step.

Step 3 — Enablement: Details That Matter

• Provide at least one complete embodiment: inputs → steps → outputs → acceptance criteria.
• Include parameter ranges, tolerances, environmental conditions, dependencies (versions, tooling).
• Add negative results and edge conditions that informed final choices.

Step 4 — Variants and Edge Cases

• Describe plausible alternatives (algorithms, materials, topologies) and trade‑offs.
• Explain fallbacks if a dependency is unavailable (e.g., storage swap, sensor alternative).

Step 5 — Diagrams and Examples

• Include labeled figures: block diagrams, flowcharts, sequence/state diagrams, mechanical drawings.
• Ensure figures are referenced in text (Fig. 1A) and align with terminology.

Step 6 — Review and Package

• Run the scoring checklist.
• Peer review with engineering → in‑house IP → counsel.
• Export clean package with attachments list and versions.

Do / Don’t

• Do: be specific, quantify, cite sources, include a reproducible path.
• Don’t: rely on marketing language, omit dependencies, or leave parameters as “TBD.”