Anatomy of a Patent Claim (and How It Gets Invalidated)
A patent isn’t invalid because a product “feels similar.” It’s invalid element by element. Below, take a real claim apart, drop prior-art references against it, and watch an invalidity claim chart build itself — the same analysis that decides petitions at the PTAB.
Select a reference to test for anticipation
Under §102, a single prior-art reference must disclose every element of the claim.
Claim 1. A self-balancing personal transport device comprising:
a platform configured to support a standing rider;
Not yet shown in the prior art
two coaxial wheels rotatably coupled to opposite sides of the platform;
Not yet shown in the prior art
at least one electric motor operatively coupled to drive the wheels;
Not yet shown in the prior art
an inertial sensor configured to detect a tilt angle of the platform; and
Not yet shown in the prior art
a controller configured to adjust motor torque based on the detected tilt angle so as to maintain the platform in an upright, balanced position.
Not yet shown in the prior art
Pick one reference. Anticipation requires a single reference to teach every element.
| Claim 1 limitation | Disclosed by | Supporting disclosure |
|---|---|---|
| A self-balancing personal transport device comprising: | ||
| [A]a platform configured to support a standing rider; | — | Not shown — this gap defeats the challenge. |
| [B]two coaxial wheels rotatably coupled to opposite sides of the platform; | — | Not shown — this gap defeats the challenge. |
| [C]at least one electric motor operatively coupled to drive the wheels; | — | Not shown — this gap defeats the challenge. |
| [D]an inertial sensor configured to detect a tilt angle of the platform; and | — | Not shown — this gap defeats the challenge. |
| [E]a controller configured to adjust motor torque based on the detected tilt angle so as to maintain the platform in an upright, balanced position. | — | Not shown — this gap defeats the challenge. |
A claim is a checklist, not a paragraph
Every patent claim is a single sentence built from limitations — the preamble plus a list of elements, here labelled [A] through [E]. That structure is the whole game: to infringe a claim a product must have every element, and to invalidate it the prior art must teach every element. Miss one and the analysis fails. The tool above makes that “all-elements rule” literal — the claim only turns fully green when every limitation is accounted for.
1§102 — Anticipation needs a single reference
Start in §102 Anticipation mode and click one reference. Anticipation is the strict ground: a single prior-art reference must disclose every element, arranged as in the claim. Try US 5,791,425 — it teaches the platform, the wheels, the tilt sensor, and the balancing controller, but it’s silent on an electric motor. Four of five isn’t anticipation; that one gap defeats the challenge. Now try US 6,302,230: it teaches all five at once, and the claim is anticipated.
2§103 — Obviousness lets you combine
Switch to §103 Obviousness and you can assemble elements from several references. Select US 5,791,425 (the balancing platform) together with US 7,000,001 (an electric skateboard, which supplies the missing motor) and the combination now covers all five elements. This is how most real invalidations work — decisive prior art rarely arrives in one tidy document.
The catch: motivation to combine. You can’t mix references with hindsight. §103 requires a reason a skilled person would have combined them — and the tool makes you acknowledge that rationale before it will call the claim obvious. That gate is exactly where many obviousness challenges are won or lost (see KSR v. Teleflex).
3The claim chart is the deliverable
As you assign references, the invalidity claim chart at the bottom fills in — each limitation paired with the reference that teaches it and the exact supporting language. This element-by-element table is the actual work product attorneys file with a petition or hand to opposing counsel. Reading one fluently is the core literacy of invalidation work, and building one is what a prior-art search is ultimately for.
Why this connects to search
A claim chart is only as strong as the references in it — and the hardest element to fill is usually the one described in unexpected words (our missing electric motor lives in a “skateboard” patent). That is precisely the gap semantic search closes: finding the reference that teaches a limitation even when it never uses the claim’s vocabulary. See the companion explainer, How AI Prior-Art Search Actually Works.
Build a real claim chart faster
PatentScan finds the references that teach each limitation — including the ones hiding behind different words — and helps you map them to the claim, element by element.