“We did it first” isn’t enough
what this meat-processing robotics patent fight teaches about oppositions
A recent Australian Patent Office decision is a useful reminder that patent oppositions are won on evidence and drafting discipline, not vibes.
In Intelligent Robotics Pty Ltd v Scott Automation & Robotics Pty Ltd [2026] APO 5 (6 February 2026), the Patent Office dealt with oppositions against two related patents for an automated carcass butchering system. The opponent ran the classic arguments—not new and not inventive—but the case ultimately turned on two practical realities:
Prior use only counts if it was actually made public, and you can prove it.
Even if you beat novelty and inventive step attacks, your patent can still be in trouble if the claims aren’t properly supported by the specification.
The invention (in plain English)
The invention is an automated carcass cutting system where carcasses hang from a conveyor. As each carcass passes a stabilising belt, an imaging system captures information, software calculates a cutting path, and a robot performs the cut.
What was being opposed?
Standard patent application 2016429338 (accepted and opposed), and
Innovation patent 2023100082 (a divisional of a divisional; certified and also opposed).
The opponent asked for the two oppositions to be joined, and the delegate allowed them to be dealt with together.
1) Novelty: “prior use” failed because the evidence didn’t show public disclosure
The opponent tried to knock the patent out by relying on prior use of a robotic carcass butchering system referred to as FACCSA v1.
The problem wasn’t the story—it was the proof.
To succeed on prior use in a patent opposition, you need to establish that the thing was made available to the public (ie, seen or used by people who were not under confidentiality obligations). On the evidence filed, the delegate wasn’t satisfied that FACCSA v1 was publicly viewed in the way the law requires.
A key witness had seen the system, but he was part of the team that invented it—so he wasn’t treated as a “member of the public” for the purpose of proving public availability.
Result: the novelty attack failed.
Business takeaway: if you’re opposing a patent and your “killer prior art” is a real-world system, you need clean evidence that it was public, when, where, and who saw it—and that there was no confidentiality restriction.
2) Inventive step: the opponent didn’t do the hard part—show the path a skilled person would actually take
The opponent also argued the invention lacked inventive step based on a combination of three prior art documents.
But inventive step isn’t just “here are three documents that mention related stuff”. You have to show why a notional skilled person would:
recognise a problem,
be led to those documents (or the relevant information),
and combine them in that way, with a reasonable expectation of success.
In this case, the delegate noted the opponent didn’t provide expert evidence to clearly explain how the skilled person would solve the technical problem, and why the three documents would be combined to arrive at the claimed system.
Result: the inventive step attack failed.
Business takeaway: oppositions on inventive step usually need proper expert evidence—not just citations. If you’re spending money opposing, don’t undercook the expert case.
3) Innovative step: barely argued, and the differences mattered
For the innovation patent, the opponent also ran lack of innovative step but provided very limited reasoning.
The delegate wasn’t persuaded the prior art documents would be treated as a single source of information, and accepted that differences between the documents and the invention amounted to a technical contribution.
Result: innovative step attack failed.
4) Support: the delegate raised it—then gave the applicant time to fix it
Here’s the part business owners often miss: even if your patent survives novelty and inventive step, it can still fall over on claim support.
The delegate exercised discretion to raise support issues and asked for submissions. In short, the delegate considered that essential features appeared to be missing from certain independent claims—particularly around:
the belt’s movement relative to the conveyor,
the belt’s inclination, and
the functional relationship between imaging and the cutting tool/path.
The applicant accepted there was an issue and indicated it was prepared to amend.
The opponent argued that even with amendments, the claims would still miss what it said was an essential feature (belt and conveyor driven at the same speed). The delegate disagreed with that specific objection, but still found the claims lacked support as filed.
Decision outcome (practically):
Claims in the standard application and innovation patent were found to lack support, and
the applicant was given two months to amend.
No costs order.
Business takeaway: “support” is where broad, ambitious claims often get clipped. If you try to claim wider than what you’ve actually described (or you leave key relationships out of the independent claims/spec), you create a target.
What this means for business owners (not patent nerds)
If you’re filing patents
Draft for reality, not marketing. If the stabilising belt and the imaging-to-cutting relationship is what makes it work, make sure the spec and independent claims clearly support those features.
Assume support will be tested later. Opponents and delegates look for missing “essential features” when claims feel broader than the disclosure.
If you’re opposing a competitor’s patent
Prior use needs hard proof of public availability—dates, witnesses, photos, records, and no confidentiality cloud.
Inventive step needs a proper expert narrative, not just document dumping.
If you’re partnering in automation (integrators, co-development)
This case is also a quiet warning: when systems are built collaboratively, the value often sits in:
configuration details,
calibration/tuning,
software logic,
and process know-how.
If you don’t lock down ownership and confidentiality in your contracts, you can end up funding someone else’s IP position.
Bottom line
This decision is a good example of how patent disputes actually play out:
“We had it first” doesn’t win unless you can prove it was public.
“It’s obvious” doesn’t win unless you can show how a skilled person would get there.
And even when the opposition fails on novelty/inventive step, support can still force amendments (or worse).
General information only, not legal advice for your specific situation.