Patent law has a talent for turning plain English into a maze with footnotes. One minute you think, “I built a better machine, so surely I can patent it,” and the next minute someone mutters “Alice step one” like it is a ghost story told in a conference room with bad coffee. Welcome to 35 U.S.C. § 101, the part of U.S. patent law that asks whether the thing you invented is even the kind of thing the patent system is willing to talk about.

For mechanical inventions, that question should sound easy. The statute literally says patents may be obtained for a new and useful process, machine, manufacture, composition of matter, or any new and useful improvement of them. A mechanical improvement sounds like it should stroll right through the front door. Often it does. But not always. In modern patent practice, the real fight is rarely over whether a claim has bolts, gears, motors, plates, shafts, or sensors. The real fight is whether the claim is directed to a specific technological improvement or whether it is drafted so broadly that it starts to look like a result, an abstraction, or a dressed-up instruction to “make the machine do the smart thing.”

That distinction matters more than ever. Recent U.S. Patent and Trademark Office guidance and a significant Federal Circuit decision involving motorized selectorized dumbbells have made one theme very clear: a mechanical improvement is on much stronger ground under § 101 when the claim recites concrete structure, explains how the improvement actually works, and ties any control logic or calculations to a real-world physical operation. In other words, patent eligibility likes machinery. It likes specificity even more.

What § 101 Really Does

Think of § 101 as the front-door test, not the whole patent exam. It does not decide whether your invention is new enough under § 102, nonobvious enough under § 103, or well described enough under § 112. It asks a narrower question: is your claimed invention the kind of subject matter patent law can protect at all?

That sounds broad because the statute is broad. It covers machines and improvements to machines. But courts have layered in three judge-made exceptions: laws of nature, natural phenomena, and abstract ideas. Those exceptions are meant to stop patents from monopolizing the basic tools of science and technology. Nobody gets a patent on gravity, the Pythagorean theorem, or the idea of “communicating information.” Nice try, though.

Courts now apply the Alice/Mayo two-step framework. First, they ask whether the claim is directed to one of those excluded ideas. If not, the claim is eligible. If yes, they ask whether the claim includes additional elements that transform the claim into a patent-eligible application. The USPTO follows that framework through MPEP § 2106 and its later updates.

Why Mechanical Inventions Usually Start Strong

Mechanical inventions usually have one major advantage in a § 101 analysis: they are often plainly about a machine doing a physical job in the physical world. That is a far better starting point than a claim directed to financial risk, data organization, or a mental process wearing a computer-shaped hat.

But a machine is not automatically safe just because it weighs more than a laptop charger. Courts have repeatedly said that merely adding generic hardware to an abstract idea is not enough. A claim can recite a controller, a transmitter, a sensor, or a processor and still fail if those components are used only as generic tools to carry out an abstract result. Physicality helps, but physicality alone does not win.

That is why patent eligibility for mechanical improvements turns on claim character. Is the claim really about a specific machine improvement, or is it about a desired outcome with vague instructions to achieve it? Does the claim recite particular components and operational relationships, or does it just announce a goal and trust the reader to fill in the engineering later? Section 101 is allergic to the second version.

The Big Cases Behind the Rule

Diehr: A mathematical equation does not poison a real industrial process

One of the most helpful Supreme Court cases for mechanical and industrial technologies is Diamond v. Diehr. The claims involved curing synthetic rubber using temperature measurements and a mathematical equation. The Court did not treat the presence of math as fatal. Instead, it looked at the claimed process as a whole and recognized that it improved a real industrial operation. That case still matters because it confirms an important principle: using calculations inside a physical process does not automatically make the process abstract. If the claim applies the calculation in a concrete technological way, it can still be patent eligible.

Chamberlain and UFRF: Generic automation can still sink a claim

Now for the caution tape. In Chamberlain, the Federal Circuit found claims ineligible where the advance was framed as wirelessly communicating status information about a system. The claims used physical components, yes, but the court saw the alleged advance as generic wireless communication using conventional technology. That was not enough. The claim lacked a specific implementation of a technological improvement.

The same general danger appeared in University of Florida Research Foundation v. GE. There, the court described the claims as automating “pen and paper methodologies” on a generic computer system. That phrase should send a chill through any patent drafter. If a claim sounds like “take the old manual process and let a general-purpose machine do it,” the § 101 risk goes up fast.

American Axle: The mechanical case that made everyone sit up straighter

If you want proof that even a hard-metal mechanical invention can collide with § 101, look at American Axle. The decision became famous because many practitioners saw it as a warning that a mechanical claim can be treated as ineligible when it effectively claims a desired result tied to a natural law without claiming the means for achieving that result with enough specificity.

That is the part inventors hate hearing. “But the product is physical!” True. Unfortunately, § 101 does not stop asking questions just because the invention can stub your toe. If a claim is drafted around a result like reducing vibration, tuning a liner, or optimizing performance, without clearly reciting the concrete way the invention gets there, the claim may be attacked as claiming the idea or principle rather than the specific implementation.

PowerBlock: A fresh win for specific mechanical improvement

Then came better news. In 2025, the Federal Circuit decided PowerBlock Holdings, Inc. v. iFit, Inc., and it gave the mechanical world something refreshing: a clear eligibility win. The court held that the claimed dumbbell system was not directed to the abstract idea of automated weight stacking. Instead, it was directed to a sufficiently specific mechanical invention.

Why did the claim survive? Because it recited actual structure and operation: an electric motor, a selector, weight plates, and the physical movement and coupling needed to automatically adjust dumbbell weight. The court emphasized that the claim was limited to a specific implementation of a technological improvement and was focused on a specific mechanical improvement. It also warned courts not to ignore conventional elements and not to conflate § 101 with novelty and obviousness under §§ 102 and 103.

That point is huge. A claim does not become abstract simply because some pieces are old. Patent law has separate tools for deciding whether a combination is new or nonobvious. Section 101 is not supposed to do that job with a fake mustache and borrowed glasses.

So What Makes a Mechanical Improvement Patent Eligible?

1. Claim the machine, not just the mission statement

If your claim reads like “a system configured to improve efficiency,” you are asking for trouble. If it reads like “a system including a motor coupled to a selector movable among defined positions to engage selected plates,” you are speaking the language that § 101 prefers. The more the claim sounds like an engineer wrote it instead of a motivational speaker, the better.

2. Show how the improvement changes operation in the real world

Claims become stronger when they do not stop at identifying components. They explain the relationship among those components and how that relationship changes machine behavior. A stronger claim does not merely say a controller receives data; it says what the controller does with the data to control heating, cooling, positioning, actuation, locking, braking, or some other physical operation in a defined way.

This is exactly why USPTO examples remain useful. In one example involving an injection molding controller, a claim becomes eligible when the measured data is used to control the molding apparatus in a particular way. The lesson is simple: using information to drive a defined physical response can help transform a claim from “just data handling” into a practical technological application.

3. If the invention uses math, sensors, or software, tie them to the physical improvement

Modern mechanical systems are often electro-mechanical systems. They use software, feedback loops, machine vision, temperature calculations, force sensing, control curves, and predictive adjustments. None of that is automatically a problem. The problem starts when the claim spotlights the calculation and buries the machinery.

Patent eligibility gets stronger when the claim reflects the technical improvement described in the specification. If the specification says your invention improves alignment accuracy, reduces vibration, prevents unsafe engagement, or stabilizes temperature in a mold, the claims should recite the features that actually produce that improvement. Not the slogan. The mechanism.

Drafting Tips That Can Save a Mechanical Application

Write the specification like the future depends on it

Because it might. The USPTO and courts look at the claims in light of the specification. A good specification explains the prior technical problem, the limits of older systems, the specific improvement, and the engineering reasons the new design performs better. That context can matter a lot when arguing that the claim is directed to an improvement in technology rather than an abstract goal.

Avoid black-box claim language

Phrases like “configured to optimize,” “adapted to determine,” or “operable to improve” are not evil, but they become risky when they replace real structure and operation. Try not to make your controller sound like a wizard in a sealed box. If the improvement depends on a particular mechanism, sensor arrangement, linkage, actuator sequence, or control rule, say so.

Do not oversell the invention into abstraction

Inventors often describe their breakthrough in broad business-like terms: safer, faster, smarter, automated, more efficient. Those words are fine for a press release. They are terrible when they become the whole legal theory of the claim. The more broadly you describe the invention, the easier it becomes for an opponent to reframe it as an abstract idea. That is exactly what happened in several bad § 101 cases.

A Simple Way to Pressure-Test a Claim

Ask this question: if someone paraphrased the claim in one sentence, would it still sound like a machine? Or would it sound like a goal?

If the paraphrase is “a system for automatically selecting weights by moving a selector with a motor through defined mechanical positions,” that sounds promising. If the paraphrase is “automating weight selection,” your claim may be in danger.

Try the same test with manufacturing equipment, medical devices, robotics, consumer hardware, or vehicle components. A claim that survives § 101 usually keeps its mechanical bones even after paraphrasing. A weak claim dissolves into a result.

What Real-World Experience Around § 101 Usually Looks Like

The lived experience around § 101 for mechanical improvements is rarely dramatic at first. It usually starts with optimism. An inventor walks into a meeting carrying a prototype, a CAD model, or a testing report and says something completely reasonable like, “We redesigned the actuator path, added feedback control, and now the device locks faster and jams less.” Everyone in the room nods. The engineering problem is real. The solution is physical. The testing data looks good. At that stage, almost nobody says, “Careful, this might sound abstract.”

Then drafting begins. That is where experience teaches its first rude lesson. The more a team tries to make the claim broad enough to cover every variation, the easier it becomes to strip out the details that made the invention persuasive in the first place. “Motor coupled to selector” turns into “adjustment mechanism.” “Defined locking position” turns into “engagement state.” “Signal to cool the mold when the measured temperature exceeds the threshold” turns into “control based on detected conditions.” Suddenly the claim is wearing a suit three sizes too big, and § 101 starts sharpening its pencils.

Another common experience is the mismatch between what engineers think is obvious from the drawings and what the legal record actually says. Engineers often assume that if a physical system is shown, everyone will understand how it works. Examiners and courts do not get to assume. They read the claim language, the specification, and the prosecution record. If the claim does not recite the feature that performs the improvement, or if the specification describes the improvement only in broad marketing language, the application can start looking less like a technological advance and more like a wish written in patent format.

Practitioners also learn that examiner conversations on § 101 often become much easier once the discussion turns concrete. Arguing in the abstract about whether something is abstract is a bad way to spend an afternoon. Arguing about a motor physically moving a selector, a controller modulating a heating element, or a mechanical linkage changing the path of force transmission is much better. Real progress tends to happen when the applicant can point to the exact claim language that captures the technological improvement and explain why the claim is not merely collecting data, calculating something, or announcing a desired result.

In litigation, the experience can be even sharper because § 101 often appears early, sometimes on a motion to dismiss. That means a patent owner may be defending the claim before getting deep into technical fact development. In that setting, claims with concrete structural detail often feel sturdier. Claims written at a high level of function can feel alarmingly fragile. That practical difference is one reason recent decisions like PowerBlock have attracted so much attention: they remind patent owners that mechanical detail still matters, and it matters early.

The most experienced teams usually come away with the same habit: they draft broad claims, yes, but they also build a ladder of narrower claims and a specification packed with technical explanation. They do not rely on the fact that the invention is mechanical. They prove it in the claim language. That is the experience-based takeaway nobody loves, but everybody remembers: under § 101, being a real machine is good; sounding unmistakably like one is even better.

Final Takeaway

A mechanical improvement can absolutely be patent eligible under § 101. In fact, it often has a strong head start. But the modern cases show that courts are not handing out automatic passes just because an invention includes hardware. Eligibility turns on whether the claim is directed to a specific technological improvement with concrete structure and operation, not merely a result dressed up in machine language.

If you remember one thing, make it this: claim the way the machine works, not just what you hope it accomplishes. That is how a mechanical invention stays a mechanical invention all the way through the § 101 analysis.