Two Examiners, One Application, Opposite Conclusions: Handling Subjectivity in Inventive Step

In my 20 years of practice as a patent attorney and patent litigator, the most uncomfortable truth I explain to clients is this: on inventive step, two competent examiners can look at the same application and the same prior art and reach opposite conclusions. That is not a scandal. It follows from what the test asks. But it does mean the applicant's job is to reduce the room for misreading.

Can two examiners really disagree about the same application?

Yes. Here is a real case.

A client filed an application for an anomaly detection method for industrial equipment. While the equipment is running normally, the method continuously updates the threshold used to judge faults. When it detects a load switch, it pauses the threshold update, and resumes it once operation has stabilized. The purpose is to stop the vibration during the switch from contaminating the baseline used for judgment.

Two prior-art references were relevant. Reference 1 disclosed dynamically updating a threshold. Reference 2 disclosed pausing fault alarms while the equipment is starting up.

Because this patent mattered a great deal to the client, I asked two examiners to conduct a mock examination.

Examiner A: Start-up and load switching both cause signal fluctuation. Since it is known that alarms can be paused, it would also be easy to think of pausing the threshold update. This is a routine combination, and the application should be rejected for lack of inventive step.

Examiner B: "Pausing the alarm" and "pausing the threshold update" do different things. The former merely stops outputting alarms for a while; abnormal data can still enter the baseline. The latter prevents the baseline from being contaminated, which affects the accuracy of judgment after normal operation resumes. The prior art gives no clear teaching toward this combination, and the patent should be granted.

The two examiners did not disagree about the facts. They disagreed about the technical role of one feature. Examiner A saw "pausing something during instability" as one general idea; Examiner B saw two different technical functions. That is where an applicant's argument has to work.

Why is inventive step so open to different readings?

Most systems use a structured test. Europe applies the problem-solution approach; China's Guidelines for Patent Examination use a comparable three-step method of identifying the closest prior art, determining the distinguishing features and the technical problem they actually solve, and judging whether the prior art as a whole would have prompted a skilled person to adopt them. The United States asks whether the claimed invention as a whole would have been obvious to a person having ordinary skill in the art.

Each version contains a judgment call: whether a teaching in one reference "suggests" its use in another context. In the case above, everything turned on whether pausing an alarm suggests pausing a baseline update. A structured test narrows the disagreement, but it cannot eliminate it.

How do I prepare an application for that uncertainty?

From cases like this one, I work in a fixed order.

1. Strengthen the concrete technical difference. Explain what each distinguishing feature actually changes, not simply that "the effect is better." In the anomaly detection case, the point is not that detection is more accurate in general, but that the baseline stays clean because contaminated data never enters it, whereas pausing the alarm leaves the baseline exposed.

2. Require the combination reasoning to have a basis. Anticipate findings that a feature is "common knowledge" or "easy to think of." Explain, ahead of time, why a skilled person would not have applied that measure to solve the problem your invention addresses. The question is not whether the two references can be put side by side, but whether anything in them points toward this particular use.

3. Amend precisely when necessary. Use timing diagrams, data flows and comparison results to show the causal link between the technical means and the technical effect. Amendments must stay within what the original application disclosed, so this evidence and explanation need to be grounded in the real R&D.

4. Use the remedies the system provides. A rejection is not the end. In China, an applicant who disagrees with a rejection decision can request reexamination; other systems provide appeals. These procedures exist because examination involves judgment.

Do not try to predict which examiner you will get. Write and argue the application so that even a skeptical examiner has to engage with the specific technical difference, rather than dismissing it as a routine combination.

When is a mock examination worth it?

For an important application, I find it worthwhile. In the case above, the mock examination surfaced Examiner A's line of reasoning, which made it clear what the explanation of the technical difference had to cover.

A mock examination does not predict the outcome of real examination. Its value is in showing you the strongest objection early, while you can still prepare for it. This is not legal advice; consult a qualified patent attorney for your situation.

What should applicants take away?

• Expect inventive step to involve judgment, and prepare for the less favorable reading.
• Describe the technical function of each distinguishing feature, not just its result.
• Challenge any combination that has no stated reason in the prior art.
• Keep diagrams, data flows and comparison results that tie means to effect.
• Treat reexamination or appeal as a normal part of the process, not a failure.

A thorough prior-art search before filing makes all of this easier, because you know which references an examiner is likely to combine and can address them from the start.