Artificial Intelligence Industry Association Inc v. Fanuc America Corp

I. Executive Summary and Procedural Information

• Parties & Counsel:
  • Plaintiff: Artificial Intelligence Industry Association, Inc. (Florida)
  • Defendant: FANUC America Corporation (Delaware)
  • Plaintiff’s Counsel: Fry Law Corporation
• Case Identification: 4:26-cv-00860, N.D. Cal., 01/27/2026
• Venue Allegations: Venue is based on Defendant maintaining regular and established places of business within the Northern District of California, specifically regional offices in Union City and Lake Forest.
• Core Dispute: Plaintiff alleges that Defendant indirectly infringes a patent related to generating synthetic image data for machine learning by promoting, integrating, and distributing robotic systems that incorporate a third-party software (Osaro SightWorks) alleged to practice the patented methods.
• Technical Context: The technology concerns the automated creation of computer-generated (synthetic) images, which are used as training data for artificial intelligence models in the field of computer vision.
• Key Procedural History: The complaint alleges that Plaintiff provided Defendant with notice of the asserted patent through correspondence beginning on or around October 27, 2025, followed by a formal demand letter on January 9, 2026. The complaint also notes that a corrective assignment for the patent was submitted to the U.S. Patent and Trademark Office on December 5, 2025.

Case Timeline

II. Technology and Patent(s)-in-Suit Analysis

U.S. Patent No. 11,257,272 - "Generating Synthetic Image Data For Machine Learning"

The Invention Explained

• Problem Addressed: The patent’s background section describes the difficulty and expense of acquiring the vast amounts of high-quality, specialized image data needed to train machine learning models for computer vision tasks. It notes that manual data collection is time-consuming and that existing datasets are often limited in size and scope (e.g., lacking stereo pairs or additional data channels like depth information). ’272 Patent, col. 1:4-51
• The Patented Solution: The invention provides systems and methods to automatically generate large-scale, richly annotated synthetic image datasets. The system programmatically assembles virtual 3D scenes by combining elements from databases of background images, 3D objects, and textures. It then uses one or more "virtual cameras" with precisely defined settings that can mimic real-world camera properties to capture images of the scene from various perspectives. This process allows for the efficient creation of diverse training data, including stereo image pairs and corresponding data channels like depth maps and segmentation masks. (’272 Patent, Abstract; col. 3:22-44; Fig. 11).
• Technical Importance: This automated approach aims to solve a fundamental bottleneck in developing computer vision AI by replacing manual, resource-intensive data collection with a scalable and customizable digital pipeline. ’272 Patent, col. 2:40-51

Key Claims at a Glance

• The complaint asserts at least independent method Claim 17. Compl. ¶¶17, 21
• The essential elements of Claim 17 include:
  • Receiving databases of background images, 3D models, texture materials, scene metadata, and camera setting files.
  • Constructing a first synthetic image scene of a specified class using a computer graphics engine, including selecting a background, populating it with 3D models, and applying textures.
  • Placing a virtual camera within the scene.
  • Incrementally varying a parameter in the camera settings file to generate a first series of unique camera views.
  • Rendering synthetic images from each of these camera views.
  • Rendering additional data channels (e.g., depth, segmentation) for the first scene.
  • Constructing a second synthetic image scene of the same class.
  • Capturing a second series of camera views of the second scene from the same positions as the first series.
  • Rendering synthetic images and additional data for the second scene.
  • Assembling the resulting synthetic images and additional data into a training dataset for a machine learning system. ’272 Patent, col. 62:1-30
• The complaint reserves the right to supplement its allegations following discovery. Compl. ¶1

III. The Accused Instrumentality

Product Identification

• The accused instrumentality is the integrated "FANUC-Osaro solution," which combines FANUC robotic hardware, such as the M10iD/12 six-axis robot arm, with Osaro, Inc.'s "SightWorks" machine-learning vision software and its "AutoModel AI engine." Compl. ¶¶1, 23, 25

Functionality and Market Context

• The complaint describes SightWorks as an "AI-driven vision system" that enables robotic systems to perform tasks such as object recognition, depth perception, and manipulation, particularly in warehouse automation and fulfillment operations. Compl. ¶¶13, 25 The core of the infringement theory rests on the allegation that Osaro uses "simulated data" or synthetic data generation techniques to create the training datasets required for its AI models, and that these techniques practice the methods claimed in the ’272 Patent. Compl. ¶25

IV. Analysis of Infringement Allegations

No probative visual evidence provided in complaint.

Claim Chart Summary

• The complaint does not provide a detailed, element-by-element mapping of the accused functionality. Instead, it makes broad allegations that Osaro’s software, upon information and belief, practices the patented methods. The following chart summarizes these narrative allegations against the elements of Claim 17.

’272 Patent Infringement Allegations

Identified Points of Contention

• Evidentiary Questions: The complaint explicitly acknowledges that "technical discovery will be necessary" to establish the specific methods used by Osaro's software. Compl. ¶1 A central question will be whether discovery reveals evidence that Osaro's use of "simulated data" practices the specific, multi-step process recited in Claim 17, including limitations for which no specific facts are currently alleged, such as "incrementally varying at least one parameter" and "constructing a second synthetic image scene."
• Scope Questions: The case is for indirect infringement against FANUC, which depends on proving direct infringement by a third party (Osaro or its customers). Compl. ¶28 The analysis will question whether the promotion and integration of a general-purpose robotic arm with third-party software rises to the level of inducing or contributing to the specific method of generating training data, which may occur separately from the robot's end-use operation.

V. Key Claim Terms for Construction

The Term: "synthetic image scene"

• Context and Importance: This term defines the virtual environment from which images are captured. Its construction is critical because it determines the necessary components and structure of the environment that must be created to infringe. Practitioners may focus on this term to dispute whether the accused method of generating simulated data involves creating a "scene" with the specific composition (backgrounds, 3D models, textures) described in the patent.
  • Evidence for a Broader Interpretation: The patent states that virtual scenes "may be assembled using a graphics rendering engine" from various components, suggesting flexibility in the implementation. Patent, col. 44:28-34
  • Evidence for a Narrower Interpretation: The specification provides detailed descriptions of assembling scenes by "arranging a background image in a background portion of the image scene and the 3D model in a foreground portion," which may suggest a required structural arrangement. Patent, col. 60:15-19

The Term: "incrementally varying at least one parameter... to provide a first series of camera views"

• Context and Importance: This term defines a key action for generating a dataset of multiple images. The dispute may turn on whether the accused system's method for creating different perspectives meets the "incrementally varying" and "series" requirements. If the accused system generates views using random or non-sequential parameter changes, it may fall outside the scope of this limitation.
  • Evidence for a Broader Interpretation: The patent describes this process as a way to "capture a variety of image scene perspectives," which could be read to cover any systematic generation of different views. Patent, col. 60:8-9
  • Evidence for a Narrower Interpretation: The specification discusses routines for "incrementally varying camera orientation" to "create hundreds or thousands of images depicting unique perspectives of a scene in minutes or seconds," which suggests a methodical, sequential process rather than a random or ad-hoc one. Patent, col. 38:46-52

VI. Other Allegations

Indirect Infringement

• The complaint alleges FANUC is liable for both induced and contributory infringement.
  • Inducement (35 U.S.C. § 271(b)): The claim is based on FANUC's alleged affirmative steps to encourage infringement, including technically integrating Osaro's software, conducting joint marketing and demonstrations, directing its network to deploy the solution, and providing public endorsements. Knowledge and intent are alleged to arise from pre-suit notice of the patent provided on or around October 27, 2025. Compl. ¶¶26, 31, 33
  • Contributory (35 U.S.C. § 271(c)): The claim is based on allegations that FANUC's robotic hardware is a "material component" of the infringing system, is "specially adapted for use with Osaro's software," and that the combined FANUC-Osaro system has "no substantial non-infringing use" in the context of their collaboration. Compl. ¶27

Willful Infringement

• Willfulness is alleged based on FANUC's continued promotion and facilitation of the accused solution after receiving Plaintiff's demand letter and offer to license. Compl. ¶39

VII. Analyst’s Conclusion: Key Questions for the Case

• A primary issue will be one of evidentiary proof: Can Plaintiff, through discovery, convert its "information and belief" allegations into concrete evidence showing that Osaro's method of generating "simulated data" practices each specific limitation of the asserted method claim, particularly the requirements for constructing multiple scenes and incrementally varying camera parameters?
• A second core issue will be one of imputed intent: For the indirect infringement claims, the case will question whether FANUC's promotion of a robotic solution with an advanced vision system demonstrates the specific intent to encourage infringement of a patented data generation method, or if its actions were merely aimed at selling general-purpose robotic hardware.
• A third question will be one of component vs. system: The contributory infringement claim will likely turn on whether FANUC’s robotic arm is viewed as a staple article of commerce with substantial non-infringing uses, or if, in the specific context of the FANUC-Osaro collaboration, it becomes a component specially adapted for an infringing system with no other substantial use.