Background
GROOVE X K.K. filed a patent application in November 2022 for a “Robot, Program and Method” (Application No. 2022-177451) with a priority date of September 2017. The claimed invention relates to a robot with an eye control unit that displays eye images on a display device with a distinctive feature: the eye control unit causes the pupil region to vibrate around a central axis point while simultaneously moving that axis point in a wave-like manner—effectively simulating human fixational microsaccades to create a more “alive” appearance.
The Japan Patent Office rejected the application in January 2024, finding it lacked inventive step under Article 29(2) of the Patent Law. After GROOVE X filed an appeal and submitted amended claims in April 2025, the Patent Office maintained its rejection decision in September 2025, concluding the claimed invention would have been obvious to a person skilled in the art by combining three prior disclosures with common technical knowledge about human eye movements.
GROOVE X then filed this action in October 2025 seeking to overturn the rejection decision, arguing that the Patent Office mischaracterized the prior art and improperly applied common knowledge about human eyes to robot eyes without sufficient motivation or evidence that such application would enhance human-robot communication.
The Court’s Holding
The Intellectual Property High Court rejected all three of GROOVE X’s arguments for invalidating the Patent Office’s decision and upheld the rejection. The court held that the claimed invention would have been obvious to a skilled artisan by combining the prior art with common technical knowledge.
Specifically, the court found that: (1) the third cited reference clearly addresses achieving fast human-like eye movements including microsaccades through projector-based display, making the Patent Office’s characterization correct; (2) the application of human eye characteristics to robot eyes for enhancing user relatability is an appropriate design choice well within the scope of choices a skilled artisan would make; and (3) the connection between fixational microsaccades and improved human-robot communication is established in the prior art, particularly in the third reference. The court found no evidence that such application would cause an “uncanny valley” effect, and the patent specification itself indicates that microsaccade-like movements enhance the robot’s biological appearance and achieve the invention’s intended purpose of making robots appear more present to users.
Key Takeaways
- Motivation to Combine: When two prior art references share a common problem-solving goal and technical approach (here, making robot eyes more human-like to enhance communication), combining them has adequate motivation under Japanese patent law, even if one reference is specifically about robot eyes and another addresses human eye movements.
- Design-Level Choices: Selecting specific features from human physiology to apply to robots—such as fixational microsaccades—falls within routine design choices available to skilled artisans in the field, provided the references establish a shared technical goal or philosophy.
- Uncanny Valley Defense Rejected: A general philosophical argument that applying human-like features might cause psychological discomfort (the “uncanny valley” phenomenon) lacks merit without concrete evidence in the record, particularly when the patent specification demonstrates that the feature achieves its intended technical effect.
- Disclosure vs. Embodiment: Prior art need not explicitly disclose every structural detail of an invention; if it discloses relevant elements and a skilled artisan would find it obvious to combine them for a known purpose, rejection for lack of inventive step is appropriate.
Why It Matters
This decision reinforces the Japanese Patent Office’s approach to evaluating robot technology and human-machine interface inventions. The court’s rejection of the “uncanny valley” defense is particularly significant, establishing that general concerns about user psychology are insufficient to overcome obviousness rejections without specific evidentiary support. This may affect future patent strategies in robotics and AI interfaces where applicants rely on arguments about user experience or perception.
The decision also clarifies that when multiple prior disclosures address the same underlying technical problem—here, making robots more relatable—their combination for that purpose is obvious, even if the individual references focus on different technical implementations (display-based eye systems vs. mechanical drive mechanisms). For applicants in the robotics field, this suggests that inventive step protection requires either novel technical combinations that are not suggested by any combination of prior art, or clear evidence that the combination would encounter unexpected technical or practical obstacles not disclosed in the prior art.