Doctrine of Repair and Reconstruction¶

A United States patent-law boundary that permits an owner or authorized user to preserve a particular patented article, while treating work that in substance makes a new patented article as infringing reconstruction.

Version: v2 · 2026-09-06 · History
Domain-specific #: 1705
Origin domain: United States patent law
Subdomain: patent exhaustion and infringement
Aliases: Repair–reconstruction doctrine, Patent repair doctrine, Permissible repair doctrine

Core Idea¶

The doctrine of repair and reconstruction is a United States patent-law boundary between two acts performed on a particular patented article. Permissible repair preserves, restores, or adapts the useful capacity of the article already lawfully acquired or otherwise authorized. Impermissible reconstruction crosses from continued use of that article into making the patented invention anew. Because 35 U.S.C. § 271(a) makes unauthorized “making” an act of infringement, reconstruction can infringe even though the actor owns old components and even though none of the replaced components is separately patented^[1]^[2].

The Supreme Court's governing formulation in Aro Manufacturing Co. v. Convertible Top Replacement Co. asks whether the patented entity, viewed as a whole, had become spent and whether the intervention in substance made a new article—a “second creation” of the claimed entity^[2]. The doctrine therefore does not classify work by ordinary-language labels such as repair, refurbishment, remanufacture, refill, upgrade, or rebuild. It compares what the actor did with the patent claims and with the continuing identity of the particular article.

The doctrine mediates two property regimes. A patent grants its owner a federal right to exclude others from making the invention. An authorized sale of a patented article, however, exhausts patent rights in that particular item and transfers it into the purchaser's personal property, carrying ordinary powers to use, maintain, repair, and resell it. Exhaustion does not grant a right to manufacture another copy. Repair/reconstruction locates the boundary: is this continued use of the sold item, or a new exercise of the patented right to make?

This is not a mechanical percentage test. Courts consider the claimed invention as a whole and the substance of the intervention. The useful lives and replaceability of parts, whether the whole was spent, the nature and extent of the work, the persistence of original components, and the ordinary expectations of maintenance can matter. No one factor automatically decides every case, and a patent owner cannot convert every unpatented component into the protected “heart” of a combination merely by emphasizing its importance.

Structural Signature¶

Patent claim defining the protected article — lawfully made and acquired or otherwise authorized specimen — owner or successor entitled to continue using that specimen — maintenance, replacement, refurbishment, or modification operation — claim-relative whole-article identity inquiry — same article preserved, or new patented article made — permissible repair or infringing reconstruction.

Claimed entity: the patent claims, not marketing terminology or a repair invoice, define the patented combination against which the work is assessed.
Authorized article: the doctrine starts with a particular patented item whose making and acquisition did not themselves infringe, or whose use is otherwise authorized. A thief or maker of an unauthorized original cannot manufacture a repair privilege from wrongful possession.
Continuing-use entitlement: sale-based patent exhaustion or another authorization frees continued use of that item from the patent owner's control. The entitlement can extend through resale to later owners and to repairers acting for them.
Intervention: parts may be cleaned, adjusted, sharpened, replaced, resealed, refilled, resized, or reassembled. The label chosen by the actor does not control.
Whole-article test: the inquiry asks whether the particular claimed article persists through the intervention or had become spent and was recreated. Component count, cost, physical size, and claimed importance are evidence, not universal thresholds.
Legal output: preservation of the same article is repair and therefore noninfringing on this ground; making the patented entity anew is reconstruction and falls within the patent owner's exclusive making right unless separately licensed.
Third-party consequence: a supplier or service firm may rely on the owner's repair privilege, but contributory or induced infringement becomes possible when the customer's underlying act is reconstruction rather than repair and statutory elements are otherwise met.

A compact representation is:

\[ J(C,A_0,W,A_1)= \begin{cases} \text{repair}, & A_1\text{ is the continued claimed article }A_0,\\ \text{reconstruction}, & W\text{ in substance makes a new instance of }C. \end{cases} \]

Here (C) is the asserted claim, (A_0) the authorized article before work, (W) the work performed, and (A_1) the resulting article. This is a diagnostic schema rather than a statutory formula or claim that adjudication can be reduced to arithmetic.

What It Is Not¶

Not patent exhaustion itself. Exhaustion terminates patent-law control over a sold item. Repair/reconstruction asks whether later conduct remains use of that item or constitutes making another patented article. The doctrines cooperate but answer different questions.
Not a general statutory right to repair. It creates no general duty for manufacturers to supply parts, manuals, passwords, diagnostic software, or tools. Warranty, contract, antitrust, trademark, trade-secret, copyright, and anti-circumvention rules can raise separate issues.
Not technical repairability. Repairability describes whether an artifact can be restored cheaply and reliably. The legal doctrine can classify an easy physical intervention as reconstruction, or a difficult, invasive intervention as repair.
Not ordinary-language remanufacturing. Commercial labels such as refurbished, rebuilt, remanufactured, reconditioned, or single-use do not decide the patent question. The legal test remains claim-relative and fact-sensitive.
Not an element-count rule. Replacing one part can be reconstruction in some circumstances; replacing several parts can remain repair in others. Courts resist a fixed percentage of original material.
Not a “gist of the invention” test for combination patents. Aro rejected treating an unpatented component as separately monopolized merely because it is important to the combination^[2].
Not permission to copy a patented article. Ownership of one specimen supports continued use of that specimen, not manufacture of additional specimens.
Not a cross-jurisdictional universal. Other legal systems may recognize repair, exhaustion, implied-license, non-derogation, or spare-parts rules with different tests and statutory settings. Those are comparative neighbors, not automatic instances of the U.S. doctrine.

Scope of Application¶

The doctrine applies in United States patent-infringement disputes concerning tangible patented machines, manufactures, and combinations that undergo maintenance, component replacement, refurbishment, modification, or repeated commercial use. It is particularly important where a durable whole contains consumable or shorter-lived unpatented parts, where an aftermarket supplier sells replacements, or where a remanufacturer collects spent or used articles and returns them to commerce.

Recurring settings include vehicle parts, machine tooling, filters, medical and industrial equipment, cartridges, single-use cameras, and other products for which the asserted patent covers a combination rather than every individual component. The doctrine can determine direct infringement by the owner or repairer and thereby affect induced or contributory-infringement claims against service providers or component suppliers.

Its application requires jurisdictional and temporal discipline. The current Patent Act grants the right to exclude others from making, using, selling, offering to sell, or importing the invention, and § 271(a) defines unauthorized making and use as infringement. Impression Products v. Lexmark confirms that an authorized sale exhausts patent rights in the item regardless of purported post-sale patent restrictions and whether the sale occurs domestically or abroad^[3]. Earlier Federal Circuit cases that imposed domestic-sale limits on exhaustion should not be imported as current territorial law, even where their repair analysis remains instructive.

The doctrine does not by itself resolve process patents, software copying, design-patent spare-parts disputes, contract restrictions, warranty coverage, safety regulation, or digital access controls. Those questions may interact with repair practice, but their governing rules are distinct. The node is an analytical map of one patent-infringement boundary, not legal advice about a particular product.

Clarity¶

The fastest diagnostic is to ask five questions in sequence:

What exactly do the asserted claims cover? A component, a combination, or the entire article?
Was this particular article lawfully made and sold or otherwise authorized? If not, exhaustion and the repair entitlement may never attach.
What failed, wore out, or was changed? Identify the component's expected relation to the useful life of the whole rather than relying on its cost alone.
What operations were actually performed? Trace removal, replacement, machining, reassembly, and addition of new structures.
Did the work preserve the same claimed article, or did it make a new instance after the old whole was spent? This is the controlling classification.

Two tempting shortcuts are unreliable. First, “most original parts remain” does not conclusively prove repair; reconstruction may reuse a substantial frame or shank. Second, “the patented feature was replaced” does not automatically prove reconstruction; in a combination claim, an unpatented short-lived component can be replaced without making the whole anew. The doctrine makes the claim-defined whole, its useful continuity, and the substance of the work visible at once.

Manages Complexity¶

Without the doctrine, every act of maintenance would sit uneasily between the patent owner's exclusive right to make and the purchaser's ordinary power to use owned property. A combination patent could otherwise be leveraged into recurring control over every replacement of an unpatented wear part. At the opposite extreme, calling every remanufacturing operation “repair” would let a purchaser convert one authorized article into an unlimited license to manufacture the invention.

The doctrine compresses that conflict into a stable legal classification. It routes work that preserves the paid-for article to the owner's use entitlement and routes work that creates another claimed article to the patentee's making right. That classification, in turn, organizes questions about component suppliers and repair services: if the end user's operation is permissible repair, selling the unpatented replacement ordinarily lacks the direct infringement predicate required for contributory infringement; if the operation is reconstruction, secondary liability becomes possible when the remaining statutory conditions are satisfied.

This compression is necessarily standards-based. Products differ in architecture, expected service life, consumables, joining methods, and claim scope. A rigid count of parts or dollars would invite design-around and fail across industries. The price of adaptability is uncertainty at the boundary, which is why reliable analysis records facts rather than substituting slogans such as “single use,” “factory sealed,” or “remanufactured.”

Abstract Reasoning¶

The doctrine supports several disciplined inferences.

Claim scope precedes physical intuition. The same physical replacement can have different legal meaning depending on what the patent claims. If a patent claims a whole combination and the replaced item is an unpatented consumable, replacement may preserve the combination. If the patent claims the replaced article itself, fabricating it can be an unauthorized making.

Expected differential life supports repair but is not a talisman. A component designed or ordinarily expected to wear out before the whole supports an inference that replacing it continues the purchased article. Yet intent, replaceability, and useful life are facts within the whole inquiry, not unilateral powers by which a patentee can dictate the result.

Exhaustion is item-specific. Authorized sale frees the particular item from patent control; it does not authorize multiplying the number of patented articles. A remanufacturing process therefore must preserve identity across the operation rather than merely reuse some old material.

Modification can remain repair. Wilbur-Ellis treated adaptation of used fish-canning machines to different can sizes as akin to repair because it enhanced the useful capacity of the existing combinations. Restoration to an exact factory state is not the doctrine's sole safe harbor.

Commercial “single-use” designation is evidence, not law. Architecture and intended use can inform the whole-article inquiry, but an owner may sometimes lawfully refurbish an item marketed for one use. Conversely, a durable-looking remnant may be only raw material for reconstructing a spent patented entity.

Proof tracks operations and provenance. A remanufacturer asserting repair should be able to show that each accused product descends from an authorized article and what steps were performed. The camera litigation illustrates that legal classification and evidentiary proof are separate requirements.

Knowledge Transfer¶

Within U.S. patent practice, the same analytical sequence transfers across products: identify the asserted claim, establish the authorized article, map expected component lives, document the intervention, and decide whether the claimed whole persists. The sequence has recurred from nineteenth-century planing-machine knives to convertible fabrics, canning-machine modifications, carbide drills, filters, cameras, and toner cartridges.

The doctrine also transfers among institutional roles. Product counsel can evaluate whether a service program risks unauthorized making. A repair business can preserve provenance and step-by-step process evidence. Patent drafters can distinguish component claims from combination claims. Courts can separate direct infringement by the owner from possible secondary liability by a supplier. Policy analysts can distinguish the existing patent defense from proposals that would affirmatively require access to repair inputs.

Transfer outside U.S. patent law is limited. The abstract conflict between preserving an owned thing and creating a new protected thing appears in copyright, design rights, and other legal fields, but their statutes and doctrines differ. The portable skeleton—adjusting overlapping entitlement boundaries—is already represented by Property Rights and Boundary. Calling a copyright restoration rule or a British spare-parts doctrine an instance of the U.S. repair/reconstruction doctrine would erase jurisdiction-specific law.

Examples¶

Worn planing-machine knives: repair¶

In Wilson v. Simpson (1850), an authorized user of a patented planing machine replaced worn cutting knives. The Supreme Court distinguished restoring a deficient part from rebuilding the combination in gross^[4]. The knives were short-lived working parts, while the machine's identity and other structure persisted. Mapped back: authorized machine → expected wear component → replacement restores operation → same claimed machine continues → permissible repair.

Convertible-top fabric: repair¶

In Aro (1961), the asserted patent covered a convertible-top combination, not the fabric alone. The fabric normally wore out years before the remaining structure. Replacing that unpatented fabric maintained the patented combination rather than making the whole anew. The replacement seller therefore lacked the direct infringement predicate for contributory infringement. Mapped back: combination claim → shorter-lived unpatented element → owner replaces element → existing combination survives → repair.

Canning-machine adaptation: repair-like modification¶

In Wilbur-Ellis (1964), purchasers modified used fish-canning machines so they could handle different can sizes. The Supreme Court treated the changes as akin to repair because they bore on the useful capacity of the old combinations for which consideration had already been paid^[5]. Mapped back: authorized old combination → adaptation of operating capacity → identity of the existing machine persists → no new claimed machine made.

Carbide drill retipping: reconstruction¶

In Sandvik v. E.J. (Federal Circuit 1997), the tip of a patented drill was brazed to its shank and not designed as an ordinary replacement part. After the tip could no longer be resharpened, the drill had reached the end of its useful life. The service removed the tip, brazed on new carbide, and machined the patented cutting geometry. Considering all facts and rejecting a bright-line rule, the court held that the process reconstructed a spent device^[6]^[n1]. Mapped back: claimed drill whole spent → new material installed → patented geometry recreated → new claimed article made → reconstruction.

Lens-fitted film packages: a process boundary¶

In Jazz Photo Corp. v. ITC (Federal Circuit 2001), reopening the shell of a used single-use camera, cleaning it, resetting the counter, installing new film and a battery, and resealing the package was permissible repair rather than reconstruction^[7]. That court also required proof that each accused specimen descended from an authorized article, and proof of the refurbishment steps actually performed on it. Fuji Photo Film Co. v. ITC (Federal Circuit 2007) reviewed the same refurbishment practices and reversed the Commission on one of the steps: replacing a camera's back cover in its entirety was part of a permissible repair rather than reconstruction, because the back had to be broken to remove the exposed film and was itself a spent part once new film was inserted^[8]. Impression Products later displaced the earlier domestic-sale limitation on patent exhaustion, so that obsolete territorial rule must not be carried forward. Mapped back: authorized specimen and process evidence → stepwise refurbishment of the existing camera body → identity of the purchased article persists → permissible repair.

Structural Tensions¶

T1: Owner autonomy versus patent exclusivity. Repair protects the purchaser's ability to use paid-for property; reconstruction protects the patentee's exclusive right to make. Expanding either side consumes the other. Diagnostic: Is the remedy preserving the purchased specimen or increasing the stock of patented articles?

T2: Flexible whole-article standard versus predictable compliance. A fact-sensitive standard works across drills, cameras, filters, and cars, but makes boundary cases costly to predict. A numerical rule would be clearer and easier to manipulate. Diagnostic: Is a proposed shortcut sensitive to claim scope and product architecture, or does it merely count parts, cost, or mass?

T3: Component importance versus combination integrity. A replaced component may supply most practical value while remaining unpatented within a combination claim. Treating it as the invention's “heart” overextends the claim; ignoring its recreation can underprotect a separately claimed component. Diagnostic: What does the asserted claim cover, as opposed to what marketing calls essential?

T4: Designed disposability versus aftermarket continuity. Manufacturers may design and label products for one use, while owners and remanufacturers discover that substantial structure survives. Design intention provides evidence about useful life but can also be used strategically to suppress aftermarkets. Diagnostic: Did the claimed whole actually become spent, or was continued use merely commercially discouraged?

T5: Restoration versus improvement. Repair sounds like return to an original state, yet Wilbur-Ellis recognizes that modification enhancing an old machine's capacity can remain within continued use. Too narrow a rule punishes adaptation; too broad a rule licenses redesign into a newly made invention. Diagnostic: Does the modification operate through the continuing article or fabricate another claimed combination?

T6: Patent permission versus practical ability to repair. A repair may be noninfringing under this doctrine yet blocked by unavailable parts, software locks, contract, warranty, safety rules, or other intellectual-property claims. Calling the doctrine a complete right to repair hides those barriers. Diagnostic: Is the issue unauthorized making under patent law, or access and liability under another regime?

Structural–Framed Character¶

Doctrine of Repair and Reconstruction is strongly framed. Its facts—what parts remain, what work occurred, and whether an artifact still functions—have physical content, but the classification exists only inside a legal institution. “Patented article,” “authorized sale,” “exhaustion,” “making,” “direct infringement,” and “contributory infringement” are jurisdiction-created roles whose significance depends on statutes and precedent.

The doctrine carries normative and distributive weight. It decides whether economic control over an aftermarket belongs to the patent owner or to owners and independent repairers. Yet its internal legal inquiry is not a policy vote about which side is more sympathetic. Courts use claim scope and article identity to administer a boundary already shaped by the Patent Act and property law.

Its vocabulary does not travel neutrally. Calling biological healing, software debugging, historical restoration, or institutional reform “permissible repair rather than reconstruction” is analogy unless an actual U.S. patent claim and authorized patented article are at issue. Its character: a historically evolved, standards-based legal frame for reconciling two entitlement bundles around one material artifact.

Structural Core vs. Domain Accent¶

What is skeletal. A portable structure remains after legal terms are stripped away: one actor has a bounded entitlement over an existing resource, another has an entitlement over reproduction of a protected pattern, and an intervention must be classified as continuation of the resource or creation of a new one. Property Rights and Boundary already carry this general architecture.

What is domain-bound. The residual requires the Patent Act's exclusive making right, claim construction, an authorized patented article, patent exhaustion or license, the whole-article identity inquiry, direct infringement under § 271(a), and possible induced or contributory liability. Its canonical distinctions arise from U.S. cases—Wilson, Aro, Wilbur-Ellis, Sandvik, Jazz Photo, and Impression Products. Remove these legal roles and the repair/reconstruction result has no controlling force.

Why this does not clear the prime bar. The node recurs across many technologies, but the substrate change is within one legal practice. A drill and a camera receive the same analysis because both are patented articles under U.S. law, not because the doctrine independently governs all physical or social systems. Its cross-domain skeleton is already cataloged; its useful residual is precisely its patent-specific test. It therefore qualifies as an autonomous domain-specific abstraction, not a prime.

Property Rights is the minimal parent the doctrine presupposes. The doctrine reconciles the patent owner's exclusive entitlement to make with the chattel owner's entitlement to use and preserve a purchased item. It demonstrates that property rights are severable and item-specific: sale exhausts the patent bundle in one specimen without conveying the separate right to manufacture additional specimens.

Boundary is related because the doctrine classifies a contested crossing between continued use and new creation. Identity and Persistence is related because adjudication tracks whether the same claim-defined article survives through component change. Maintenance describes the portable intervention that preserves function, while Repairability is a domain-specific engineering neighbor concerning how feasible restoration is. None of these nodes supplies the controlling patent-law rule, and adding them all as parents would obscure the legal entitlement boundary.

Relationships to Other Abstractions¶

Current abstraction Doctrine of Repair and Reconstruction Domain-specific

Parents (1) — more general patterns this builds on

Doctrine of Repair and Reconstruction presupposes Property Rights Prime

Property Rights is the minimal parent the doctrine presupposes.

Hierarchy path (1) — routes to 1 parentless root

Doctrine of Repair and Reconstruction → Property Rights → Boundary

Neighborhood in Abstraction Space¶

Doctrine of Repair and Reconstruction sits in a sparse region of the domain-specific corpus (99^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (1565 abstractions)

Nearest neighbors

Personal Property — 0.74
Doctrine of inherency — 0.74
Anonymous Sourcing — 0.73
Documented Plurality of Users — 0.73
Ad Verecundiam (Appeal to Authority) — 0.73

Computed from structural-signature embeddings · 2026-09-08

Not to Be Confused With¶

Patent exhaustion. Ends patent control over a particular sold item. Tell: Is the question whether rights remain in the item after sale, or whether later work made a different patented item?
Implied license. Authorization inferred from conduct or sale circumstances. Tell: Is the issue the scope of permission granted, or the repair/reconstruction classification after authority exists?
Right-to-repair legislation or policy. May require access to parts, tools, manuals, diagnostics, or software. Tell: Is an affirmative access duty sought, or only a defense to patent infringement for work on an owned article?
Repairability. An engineering property measuring feasible restoration. Tell: Is the question technical/economic ease, or whether the work counts legally as making?
Warranty law. Allocates who must pay for defects and regulates warranty conditions. Tell: Can the repair be lawful yet outside warranty coverage? If yes, the doctrines differ.
Patent misuse or antitrust. Addresses improper patent leverage or anticompetitive conduct. Tell: Is the claim about the character of the work on one article, or about market conduct and enforcement?
Copyright repair and anti-circumvention exceptions. Govern copies, software, and access controls under different statutes. Tell: Is the asserted exclusive right a patent claim's making right?
Reissue or claim reconstruction. Patent-prosecution and claim-interpretation terminology unrelated to physically reconstructing a patented article.
Ordinary refurbishment labels. Remanufactured, reconditioned, rebuilt, refilled, and single-use are evidence-bearing descriptions, not legal verdicts.

Notes¶

[n1] Checked against the reported decision, Sandvik Aktiebolag v. E.J. Co., 121 F.3d 669 (Fed. Cir. 1997), rather than against a secondary summary: the facts given above are the court's own, and nothing in the example is embroidered. ↩

References¶

[1] 35 U.S.C. § 271, Infringement of patent. Supplies the statutory rule that unauthorized making of a patented invention is an act of infringement. The two qualifications that follow it — that reconstruction can infringe although the actor owns the old components, and although none of the replaced components is separately patented — come from Aro, cited with it on the same sentence. registry ↩

[2] Aro Manufacturing Co. v. Convertible Top Replacement Co., 365 U.S. 336 (1961). Supplies the two qualifications on the § 271(a) rule: that reconstruction can infringe although the actor owns the old components, and although none of the replaced components is separately patented, because what the claim protects is the combination rather than any single element. Supplies Aro's whole-entity, 'second creation of the patented entity' test for distinguishing permissible repair from impermissible reconstruction. Supplies Aro's rejection of an 'essential element'/'heart of the invention' theory for combination patents. registry ↩a ↩b ↩c

[3] Impression Products, Inc. v. Lexmark International, Inc., 581 U.S. 360 (2017). Supplies Impression Products' holding that an authorized sale exhausts patent rights regardless of post-sale restrictions or whether the sale was domestic or foreign. registry ↩

[4] Wilson v. Simpson, 50 U.S. (9 How.) 109 (1850). Supplies Wilson v. Simpson's distinction between restoring a worn part and rebuilding the patented combination in gross. registry ↩

[5] Wilbur-Ellis Co. v. Kuther, 377 U.S. 422 (1964). Supplies Wilbur-Ellis's holding that adapting a used combination's operating capacity, once payment has been made for that combination, is akin to repair. registry ↩

[6] Sandvik Aktiebolag v. E.J. Co., 121 F.3d 669 (Fed. Cir. 1997). Supplies Sandvik's totality-of-the-circumstances holding that retipping a spent, non-resharpenable drill by brazing on new carbide and remachining the patented cutting geometry is reconstruction, not repair. registry ↩

[7] Jazz Photo Corp. v. International Trade Commission, 264 F.3d 1094 (Fed. Cir. 2001). Supplies Jazz Photo's holding that reopening a used single-use camera, cleaning it, resetting the counter, replacing film and battery, and resealing the package are permissible repair rather than reconstruction, together with its insistence on proof that each specimen descends from an authorized article; the full-back replacement step is carried by Fuji Photo, cited on the sentence that follows it. registry ↩

[8] Fuji Photo Film Co. v. International Trade Commission, 474 F.3d 1281 (Fed. Cir. 2007). Federal Circuit opinion, No. 04-1618. Supplies Fuji Photo's holding that the Commission erred in treating cameras whose full backs were replaced as impermissibly reconstructed, and that replacing the full back was instead part of a permissible repair, because the back covers had to be broken to remove the film and were spent parts that could no longer serve their function once new film was inserted; it is cited alongside Jazz Photo on the lens-fitted film package example. registry ↩