What does a patent claim chart look like when product evidence supports some limitations and leaves others unresolved? A product specification can name a processor, wireless connection, and display without explaining how they work together. A worked chart makes those distinctions visible, showing where public documentation helps, where an inference begins, and which implementation facts remain unproved.

Educational example using an invented claim and fictional product; attorneys retain review and decision authority.

What a patent claim chart records

A patent claim chart places each claim limitation alongside the evidence proposed to address it. A limitation is a requirement expressed in the claim, including relationships or operating conditions as well as named components.

For U.S. literal-infringement analysis, every limitation of the properly construed claim must be present. A collection of similar product attributes does not answer that question. The Federal Circuit explains the distinction between interpreting a claim and comparing it with a product in Union Carbide v. Shell.

The practitioner’s work includes reviewing the claim breakdown, proposed mappings, and underlying evidence. The example here concentrates on product evidence, with the unresolved rows left visible.

Patent claim chart example: a fictional status display

SignalBoard v1 is an invented desktop status-display concept. For this exercise only, assume its proposed hardware incorporates an unmodified Arduino UNO R4 WiFi board. That incorporation is a scenario assumption, not a fact established by the public documents.

The evidence consists of public documentation for the board. It does not describe SignalBoard firmware, establish an assembled product’s behavior, or document sales or use. This boundary allows hardware disclosures to be examined without inventing implementation evidence.

Complete hypothetical claim 1

The following claim is written solely for this example. It is not taken from a patent, has no patent number or filing date, and has no associated specification or prosecution history.

1. A status-display device comprising:

(a) a microcontroller;

(b) a Wi-Fi communication circuit coupled to the microcontroller;

(c) a 12-by-8 light-emitting-diode matrix controlled by the microcontroller;

(d) instructions that cause the microcontroller, upon receiving a status message through the Wi-Fi communication circuit, to select and display on the matrix a symbol representing the message; and

(e) nonvolatile memory in which the microcontroller records the received message and its receipt time before displaying that symbol.

Light-emitting diodes (LEDs) form the matrix in limitation (c). The manual calls the processor a microcontroller unit (MCU). Electrically erasable programmable read-only memory (EEPROM) is a form of nonvolatile storage. Nonvolatile memory retains information without continuous power. Limitation (e) adds particular content and timing requirements to the memory requirement.

Evidence boundaries and dates

Access dates use America/New_York (EDT, UTC−04:00). Only these 2 documents belong to the example’s source set:

• D1, hardware overview: Arduino’s UNO R4 WiFi overview. No publication date or revision identifier is displayed. Accessed October 8, 2026. Its evidence date is the access date, not a demonstrated date of historical availability or actual product use.
• D2, user manual: Arduino’s ABX00087 user manual. The latest listed change-log entry is revision 8, dated October 29, 2025, in §22, printed p. 25. The retrieved PDF separately prints the field “Modified: 08/10/2026” in its footer. Accessed October 8, 2026. Neither date establishes when SignalBoard existed or operated. All page locators below are printed PDF page numbers.

Excerpts from Arduino documentation (docs.arduino.cc), licensed under CC BY-SA 4.0.

The quoted passages below preserve the sources’ wording. Component-supported means the documentation supports the proposed board-level hardware mapping under the scenario assumption. Partial means only part of the limitation is documented. Not shown means no responsive passage was found in D1 or D2. None of these labels means a claim limitation has been legally established in a real product.

Complete limitation-by-limitation chart

The last 2 rows expose the main evidence gap. A wireless circuit, programmable matrix, and memory can exist on the same board while the claimed sequence remains undocumented. Filling those cells with affirmative descriptions would replace missing evidence with assumptions.

Mistakes this example makes visible

• Stopping at nouns. Rows (b) and (c) include connection and control requirements. Rows (d) and (e) add message semantics, causation, and ordering.
• Changing the controller between rows. Evidence about one processor cannot silently stand in for another when the claim refers back to “the microcontroller.”
• Turning capability into implemented behavior. Available memory does not establish that particular data is stored. A display’s programmability does not establish its response to a network message.
• Treating a retrieval date as a historical fact. October 8, 2026 records access to these documents. It is not SignalBoard’s release date, an actual-use date, or proof of earlier public availability.
• Treating silence as absence. “Not shown in D1 or D2” describes this source set. It does not establish that a capability cannot exist.

The next evidence request follows from the row: product-specific build records for incorporation, controller and circuit evidence for relationships, and authorized firmware or operation evidence for the sequence. Additional pages describing generic board capabilities would leave the behavioral gaps intact.

Keep product, prior-art, and standards mapping separate

Similar table layouts can answer different patent questions. The identity and role of the right-hand evidence matter.

SignalBoard’s chart is a product-evidence exercise. It makes no prior-art theory, has no patent priority date, and charts no standard clauses. The separate evidence-of-use guide explains the broader product-evidence workflow without changing those boundaries.

What attorney review adds to the chart

In an actual matter, construction comes from the patent record and applicable law. Phillips v. AWH addresses claim language in its technical and intrinsic context, including the specification and prosecution history. The alternative readings above illustrate evidence consequences; they are not adopted constructions for an actual patent.

An attorney review record should address:

1. Claim completeness: the exact patent and claim version, all limiting language, and any incorporated limitations of a dependent claim.
2. Construction: consistent meanings for recurring terms, disputed relationships, and conditions. The hypothetical claim has no intrinsic record from which to resolve these questions.
3. Product identity: model, hardware revision, firmware version, configuration, and the relationship between component documentation and the identified product.
4. Evidence integrity: complete source context, saved document or capture, locator, access date, authenticity, and any evidentiary or admissibility issue relevant to the proceeding.
5. Unresolved theories and facts: every gap, competing reading, any separately developed doctrine-of-equivalents theory, and legally relevant activity, dates, and geography.

A missing row remains an evidence task. Neither a confidence label nor the number of citations resolves it.

Use AI assistance for an inspectable draft

AI can assist with organizing limitations and proposing evidence mappings. The American Bar Association’s Formal Opinion 512, §II.A, addresses lawyers’ competence duties when using generative AI, including review of its outputs.

For a chart like SignalBoard’s, the review requirement stays concrete: preserve exact claim language, inspect each cited passage in context, retain source and version information, and leave unsupported behavior visible.

Where Patlytics fits

Patlytics’ Claim Charting module uses AI to generate limitation-level infringement and invalidity charts with supporting citations. Attorneys can edit the claim construction and re-read the chart under a different interpretation, such as the direct-coupling reading of “coupled” in row (b).

The editable charts can be exported to Word or Excel for review of claim language, mappings, and sources. Attorneys retain review and decision authority over the chart and its analysis.

See how Patlytics builds a limitation-level chart with citations traceable to their sources. Book a demo.