Looking forward to patenting your invention? The predicament of getting a rejection always lingers. Often, it may be possible to overcome the rejection, but it unnecessarily delays the allowance. Do you know what the most common reason for rejection is? – It’s § 103 or obviousness.

Did you know: § 103-type rejection accounts for the majority of all patent rejections at the USPTO. It is not only common, it is expected.

So what is a 103-type rejection?

In simple terms, you receive one when the examiner finds more than one document that jointly represents your invention as an “obvious improvement.”

Let’s say the idea is – “An unmanned drone for fighting forest fires that uses canisters filled with dry ice as fire extinguishing material. ”

Now if there exist two prior art documents:

i) One that describes the use of unmanned aerial vehicles to fight forest fires.
ii) And another that describes the use of powdered carbon dioxide (dry ice) to extinguish the fire.

Then, the idea shall be deemed obvious. You want to avoid this instance lest you want to spend hundreds of dollars in filing fees and face rejections. So what is the solution? A patentability search can help.

At the same time, practitioners often point out that 103 type rejections are not always straightforward. The rationale for combining references can sometimes feel subjective or thin, and reactions to this vary widely. Some applicants view it as a natural part of the process, while others see it as an area that needs more consistency. What everyone agrees on is that obviousness rejections are common, costly, and time-consuming to overcome.

This makes early preparation critical. By running a combinational prior art search before filing, you can surface the very kinds of reference combinations examiners are likely to cite. Instead of being surprised at prosecution, you can strengthen your claims, refine your strategy, and reduce the chance of delays. PQAI helps you do exactly that.

How to Find Combinational prior art to avoid 103 type rejection?

Let’s run this example through PQAI, our open-source search engine that helps identify combinational prior art.

We used the following idea query: An unmanned drone for fighting forest fires that uses canisters filled with dry ice as fire extinguishing material.

When we ran this query in PQAI, the engine returned 91 result combinations. Among them, two stood out as a clear illustration of how examiners might form a § 103 rejection:

The US11213706B2 patent discloses a fire extinguishing device comprising liquid nitrogen-filled canisters. These canisters are designed to be rapidly dispersed using an explosive mechanism and can be delivered by aerial systems, including drones, vehicles, or robotic platforms.

Whereas the KR102597610B1 patent introduces a drone system equipped with fire-extinguishing liquid injection technology. The drone can approach fire scenes remotely and spray liquid directly at the fire site.

Taken together, these two references demonstrate exactly the kind of “obvious combination” an examiner could use to reject our query under § 103. One patent provides the specialized canister-based extinguishing mechanism, while the other provides the drone platform for aerial delivery. Combined, they align with the essence of our proposed invention.

This example shows why running a combinational prior art search before filing is so important. By surfacing results like these early, PQAI allows inventors and attorneys to fine-tune their invention, refine claim language, and avoid costly surprises during prosecution.

What Causes § 103-Type Rejections?

A 103 type rejection, also called an “obviousness rejection”, occurs when the examiner finds that your invention is not sufficiently inventive over what already exists. Importantly, this doesn’t require a single reference to match your idea word-for-word. Instead, the examiner may combine two or more prior art references and argue that together they make your invention an obvious improvement.

The USPTO applies three main tests when issuing a § 103 rejection:

• Combination of references: It is enough if multiple references, taken together, cover all elements of your claimed invention.
• Obvious to PHOSITA: The invention must not appear obvious to a Person Having Ordinary Skill in the Art (PHOSITA) at the time of filing.
• Reason to combine: There must be an articulated rationale for why the references could reasonably be combined to achieve the claimed invention.

This approach was cemented by the U.S. Supreme Court in KSR v. Teleflex (2007), which made it clear that examiners have flexibility to use logical reasoning, not just explicit teachings, when combining references.

For applicants, this means that even if no single prior art looks like your invention, examiners can still combine references to build an obviousness case. This is why § 103 is the most common reason for rejection at the USPTO.

How to Rule Out § 103-Type Rejections with PQAI?

The most effective way to lower the risk of a § 103 rejection is to anticipate the prior art combinations an examiner might rely on. PQAI enables this by combining semantic understanding with AI-powered search.

Instead of relying on keywords, it interprets your invention in plain language and maps it to relevant patents and technical literature worldwide. It then highlights references that could reasonably be combined, giving you early visibility into the same logic an examiner might apply.

As an open source project, PQAI is continually improved by a community of inventors, researchers, and practitioners, ensuring the tool evolves with the challenges of patent searching.

PQAI in Action: Proof of Concept

To demonstrate PQAI’s effectiveness, we validated its results against real-world patent prosecutions. We selected a set of applications that had received § 103 rejections and tested whether PQAI could surface the same prior art references. In every case, PQAI identified at least one of the examiner-cited references as part of its combinational results.

Application Number	Publication Number	Priority Date	Assignee	Rejection Date*	Rejection Type	Publication No. Prior Art	PQAI Input
16246472	US10731375B2	7/6/16	NABORS DRILLING TECH USA INC	23/4/20	103	US9027287B2	Abstract
15866107	US10696381B2	9/1/18	THE BOEING COMPANY,CHICAGO,IL,US	15/4/20	103	US5908174A	Claim 1
16451553	US10754607B2	26/9/18	QUALCOMM INCORPORATED,SAN DIEGO,CA,US	16/4/20	103	US6396329B1	Abstract
15712616	US10762997B2	12/10/16	KOREA ATOMIC ENERGY RESEARCH INSTITUTE,DAEJEON,KR	22/4/20	103	US20140205052A1	Claim 1
14594517	US10777098B1	12/1/15	RAYLYNN PRODUCTS LLC,GROVE CITY,OH,US	22/4/20	103	US5121745A	Abstract
16123902	US10778561B2	8/9/17	BROCADE COMMUNICATIONS SYSTEMS LLC,SAN JOSE,CA,US	23/4/20	103	US20130266307A1	Abstract

*Rejection dates are from USPTO prosecution records.

Below is a sample of two representative cases where PQAI mirrored examiner reasoning:

Case 1: US10731375 – Side saddle slingshot drilling rig

We picked up the abstract of the subject patent application US10731375 and ran it through PQAI.

PQAI spotted a U.S. patent titled Fast transportable drilling rig system (US9027287B2) as one of the prior art references in its combinational results. This was also cited by the examiner in the rejection.

Case 2: US10696381B2 – Hydraulic systems for shrinking landing gear

We picked up the claim for this application and ran it through PQAI. 

Sidenote: When looking for prior art using PQAI for a particular patent, it’s best advised to put the invention query from any of these aspects in this order (along with the priority date filter): 

1. Abstract 
2. Independent claims one at a time 
3. Summary 
4. Embodiments from specifications

PQAI surfaced US5908174A as one of the results in its combinations. This same reference was also cited by the examiner to reject the application.

These examples show that PQAI can independently surface prior art references examiners actually relied on in type 103 rejections.

PQAI: Your Early Warning System Against §103 Rejections

103 type rejections, based on combinational prior art, are among the most common barriers in the patenting process. But PQAI turns the table by letting you test your invention before they do.

Powered by AI-driven semantic mapping, PQAI goes beyond simple keyword matching to understand concepts and relationships.

Here is what sets PQAI apart from existing tools in the market:

• Access to more than 11 million patent records across major offices and 11.5 million research papers and technical publications.
• Allow search in natural language queries that accept invention abstracts, problem statements, or plain English descriptions.
• Semantic search that identifies non-obvious connections across technologies, surfacing combinations traditional tools miss.
• Open-source architecture that grows stronger with contributions from the global innovation community.

PQAI is not just a search engine. It is an early warning system for innovators, attorneys, and R&D teams to reduce risk, refine claims, and file with confidence.

Running your idea through PQAI may be the smartest step you take before filing your patent application. What’s the wait for? Try the tool today!