A sample design-around report, in the real format
Demo notice: the product on this page is fictional (a robot-vacuum anti-tangle brush); all cited patents are real and verifiable by number; the deep-dive sections use only patents estimated to be expired (verify legal status against the official register) and are shown purely to demonstrate the report format. Nothing on this page is an analysis of any in-force patent.
What was entered
"An anti-tangle main-brush system for a robot vacuum: quick-release hybrid brush + motor-current sensing to detect tangle load + brush reversal with a comb scraper to eject hair when resistance exceeds a threshold. We want to design around the competitor's 'tangle detection + auto-reverse' approach." — one paragraph of product intent, plus one competitor patent number.
Step 1: free related-patent scan (a factual list)
The scan lists facts only: patent number, technical relevance, estimated legal status — no risk rating. A real report lists 8-10 hits; three are illustrated here, with expired ones labeled (expired patents' teachings are public domain).
- US·DEMO·0000000·B2Placeholder (not a real patent)Relevance 9/10
- …(a real report lists real, verifiable patent numbers from live search; the placeholder here only demonstrates the report format)
The 30-second view: three candidate directions
Each direction in plain words: what changes, how hard, and what the signal tier is.
Direction A: sense load via speed differential instead of motor current
Use a Hall sensor to compare target vs actual brush RPM to detect tangling — a different mechanism, a different path.
Change scope: medium (sensor + firmware)
Equivalence-risk signal: lower (not clearance; counsel review required)
Direction B: untangle via axial vibration instead of reverse + scraper
Floating end caps let the brush vibrate axially in short bursts, pushing hair into end collectors — no reversal, no comb scraper.
Change scope: large (structural redesign)
Equivalence-risk signal: lower (not clearance; counsel review required)
Direction C: keep the reversal, make the trigger manual
Drop automatic detection; a one-tap App mode triggers untangling — removes the 'automatic' element, at a UX cost.
Change scope: small
Equivalence-risk signal: borderline (demonstrated honestly: we tell you when we can't call it — which is exactly why counsel reviews before launch)
The evidence layer — for your engineers and your counsel
The breakdown below uses the placeholder number US·DEMO·0000000·B2 (illustrative, not a real patent). A real report generates this for every target patent.
Claim element breakdown (excerpt)
| Element | Claim language (illustrative) | Note |
|---|---|---|
| 1a | an autonomous cleaning robot comprising a chassis… | Generic preamble — not a target |
| 1c | a brush assembly rotatably mounted… | Structural element, limited substitution space |
| 1e | a control system configured to reverse said brush in response to a detected load condition | ★ Weakest limitation: condition + action double qualifier — the largest design space |
Deterministic scoring flags 1e as the weakest limitation: it constrains both the trigger (detected load) and the action (reversal) — removing or replacing either half escapes the claim's literal scope (All-Elements Rule). That is precisely what directions A/B/C each attack.
Doctrine-of-equivalents self-check demo — Function · Way · Result
Escaping the literal scope is not the end of the story: the doctrine of equivalents exists precisely to catch substitutions. So every candidate is fed back into the DOE engine for an adversarial test. Here is why Direction C is judged borderline:
| Claim | Substitute | Assessment | |
|---|---|---|---|
| Function | Automatically clears brush tangles | Clears brush tangles (manually triggered) | Substantially the same |
| Way | Detect load → auto-reverse | User judgment → App-triggered reverse | Different trigger, same action |
| Result | Cleaning resumes without intervention | User intervention required | Different |
Two of the three prongs are half-same — a textbook borderline case. We don't dress it up: borderline candidates are explicitly flagged 'counsel review required', never packaged as safe options.
This sample — like every real report — is an engineering design reference with ranking signals, not a legal opinion on infringement and not a formal FTO. Before making, using or selling any candidate design, commission a formal FTO from registered patent counsel (comprehensive search + claim construction + element-by-element analysis including equivalents).