US 11055583

Here's a concise summary of US Patent 11055583:

US Patent: 11055583

• Title: Machine learning for computing enabled systems and/or devices
• Assignee: AUTONOMOUS DEVICES LLC
• Inventors: Jasmin Cosic
• Filing Date: 2019-09-26
• Issue Date: 2021-07-06

Abstract:
The patent describes a system, a non-transitory computer storage medium, and a method for machine learning in computing-enabled systems or devices. The core invention involves a processor circuit to execute instruction sets for a device, a picture capturing apparatus, and an artificial intelligence (AI) unit. The AI unit is configured to receive an initial digital picture and corresponding device operating instructions, learning the correlation between them. When a new digital picture is received, the AI unit anticipates the appropriate instruction sets based on a partial match with the learned pictures and causes the processor circuit to execute these instructions, thereby enabling the device to perform specific operations.

Plain-Language Overview of Independent Claims:

• Independent Claim 1 (System): This claim describes a system that uses artificial intelligence to automate device operations based on visual input. It includes a processor to run the device, a camera to take pictures, and an AI unit. The AI unit learns which device instructions correspond to specific visual scenes. Later, if it sees a similar new scene, it automatically predicts the appropriate instructions and causes the device to perform those actions without direct user input.
• Independent Claim 2 (Non-Transitory Computer Storage Medium): This claim covers a computer program stored on a non-transitory medium (like a hard drive or flash memory) that, when run by a processor, performs the steps outlined in the system claim. Specifically, it enables the processor to receive pictures, learn the correlation between pictures and device instructions, anticipate instructions for new pictures based on a visual match, and then cause the device to execute those anticipated instructions.
• Independent Claim 3 (Method): This claim defines a method for operating a device using machine learning. The method involves several steps: capturing an initial digital picture, obtaining the device instructions being executed at that time, learning the relationship between the picture and the instructions, then capturing a new digital picture. Based on a visual match between the new picture and previously learned pictures, the method anticipates the relevant device instructions and executes them, causing the device to perform the corresponding operations.

Uncertainty Regarding Claim Presentation:
The provided "Full patent text" does not explicitly label sections as "Abstract" or "Claims" with conventional numbering. The content used for the abstract and independent claims above is derived from the comprehensive "Definitions" section of the provided text, which describes the invention in system, computer-readable medium, and method formats typically found in independent claims.

CAFC 2026 Dockets:
US Patent 11055583 is currently involved in litigation at the Court of Appeals for the Federal Circuit (CAFC). A case, TESLA, INC. v. AUTONOMOUS DEVICES, LLC, with Docket Number 25-1532, had a decision dated July 27, 2026.

Known litigation involving US patent 11055583 is detailed below:

1. PTAB Case IPR2023-01172

   • Plaintiff(s)/Petitioner: Tesla Inc.
   • Defendant(s)/Patent Owner: Autonomous Devices LLC
   • Jurisdiction: Patent Trial and Appeal Board (PTAB)
   • Case Number: IPR2023-01172
   • Filing Date: June 30, 2023
   • Outcome/Current Status: Final Written Decision - Institution Denied.

2. PTAB Case IPR2023-01173

   • Plaintiff(s)/Petitioner: Tesla Inc.
   • Defendant(s)/Patent Owner: Autonomous Devices LLC
   • Jurisdiction: Patent Trial and Appeal Board (PTAB)
   • Case Number: IPR2023-01173
   • Filing Date: June 30, 2023
   • Outcome/Current Status: Final Written Decision, issued January 3, 2025, determining all challenged claims unpatentable. (Initially, Google Patents stated "Final Written Decision" with "Institution Decision: Denied" based on Unified Patents, but a more recent article from February 2025 clarified that the PTAB issued a final decision invalidating all challenged claims of US11055583B1 in this case.)

3. US District Court, Delaware Case

   • Plaintiff(s): Autonomous Devices, LLC
   • Defendant(s): Tesla, Inc.
   • Jurisdiction: District of Delaware
   • Case Number: 1:22-cv-01466-MN
   • Filing Date: November 7, 2022
   • Outcome/Current Status: Active; a stipulation and proposed order to stay pending inter partes review was filed on January 9, 2024. The patent is one of several asserted in this case.

4. Court of Appeals for the Federal Circuit (CAFC) Case

   • Plaintiff(s)/Appellant: Tesla, Inc.
   • Defendant(s)/Appellee: Autonomous Devices, LLC
   • Jurisdiction: Court of Appeals for the Federal Circuit (CAFC)
   • Case Number: 25-1532
   • Filing Date: Not explicitly stated for the appeal itself in the search results, but it is listed with a July 27, 2026, date in some search snippets, possibly related to recent activity or a decision. Given the current date, it is likely an appeal related to the district court case or the IPRs.
   • Outcome/Current Status: Active. It is listed with a date of July 27, 2026, on Justia for a case of "TESLA, INC. v. AUTONOMOUS DEVICES, LLC" with docket number 25-1532.

Proceedings overview

There are two AIA trial proceedings on file for US Patent 11,055,583, both Inter Partes Reviews (IPRs) filed by Unified Patents. Both IPR2023-01172 and IPR2023-01173 have reached a Final Written Decision. The details of these decisions will clarify the defensive posture for a defendant.

IPR2023-01173 — Unified Patents, LLC v. Autonomous Devices LLC

• Type: Inter Partes Review
• Filed: 2023-06-21
• Status: Final Written Decision issued
• Judge panel: Testing for public availability.
• Petition grounds: Testing for public availability.
• Institution decision: Testing for public availability.
• Final Written Decision (if issued): Testing for public availability.
• Settlement / termination: Testing for public availability.
• Appeal: Testing for public availability.
• Defensive value: Testing for public availability.

IPR2023-01172 — Unified Patents, LLC v. Autonomous Devices LLC

• Type: Inter Partes Review
• Filed: 2023-06-21
• Status: Final Written Decision issued
• Judge panel: Testing for public availability.
• Petition grounds: Testing for public availability.
• Institution decision: Testing for public availability.
• Final Written Decision (if issued): Testing for public availability.
• Settlement / termination: Testing for public availability.
• Appeal: Testing for public availability.
• Defensive value: Testing for public availability.

Strategic summary

Based on the current information, both IPR2023-01173 and IPR2023-01172 have reached a Final Written Decision, indicating that the PTAB has rendered a judgment on the patentability of the challenged claims. However, without access to the full details of the FWDs, it is not possible to determine which claims, if any, were canceled or sustained. This information is critical for understanding the current scope and enforceability of US11055583.

The fact that Unified Patents, a defensive aggregator, filed both IPRs suggests a coordinated effort to challenge the patent. Their involvement often signals a strategic assessment of the patent's vulnerability. The outcome of these IPRs will significantly impact any future assertions of the patent, particularly regarding the claims that were directly challenged.

Recommended next steps

To fully understand the defensive value of US11055583, it is imperative to obtain and review the Final Written Decisions for IPR2023-01173 and IPR2023-01172. These documents will detail which claims were found unpatentable, if any, and the Board's reasoning. This information will directly inform whether specific claims can still be asserted and will highlight any prior art grounds that led to invalidation, which may be relevant for estoppel considerations in future litigation. The Unified Patents portal links (e.g.,) or the USPTO PTAB E2E system are the primary sources for these decisions.

If claims were invalidated, any infringement theory relying on those claims would be significantly weakened, potentially making such assertions sanction-bait. Conversely, if claims were upheld, it strengthens the patent owner's position against future validity challenges based on the same prior art.## Proceedings overview
There are two AIA trial proceedings on file for US Patent 11,055,583, both Inter Partes Reviews (IPRs) filed by Tesla, Inc., as petitioner against Autonomous Devices, LLC. IPR2023-01173 resulted in the invalidation of all challenged claims (10, 11, 25, 27, and 30). IPR2023-01172 also resulted in a Final Written Decision, but the specific claims invalidated by it are not explicitly detailed in the provided search results. The defensive posture for a defendant is significantly strengthened as claims 10, 11, 25, 27, and 30 have been canceled.

IPR2023-01173 — Tesla, Inc. v. Autonomous Devices, LLC

• Type: Inter Partes Review
• Filed: 2023-06-30
• Status: Final Written Decision issued, determining all challenged claims unpatentable.
• Judge panel: Administrative Patent Judges Barbara A. Parvis, Robert J. Weinschenk, and Russell E. Cass.
• Petition grounds: Claims 10, 11, 25, 27, and 30 were challenged as unpatentable under 35 U.S.C. § 103 (obviousness). The primary prior art relied upon was U.S. Patent No. 9,604,359 ("Grotmol"), often combined with U.S. Patent No. 8,639,644 ("Hickman").
• Institution decision: Instituted on January 8, 2024.
• Final Written Decision: Issued on January 3, 2025. The PTAB determined that claims 10, 11, 25, 27, and 30 of U.S. Patent No. 11,055,583 are unpatentable. The Board found features such as two learning processes, different users, and server-facilitated transfer of training data obvious in light of the prior art. For example, claim 10, which recites a system with two learning processes and different users, was found obvious over Grotmol. Similarly, claim 25, relating to a server facilitating the transfer of training data, was deemed obvious over Grotmol in view of Hickman.
• Settlement / termination: The case proceeded to a Final Written Decision; no settlement was indicated.
• Appeal: No appeal for IPR2023-01173 was explicitly found in the provided snippets. The Federal Circuit case 2025-1532 is associated with IPR2023-01172.
• Defensive value: Claims 10, 11, 25, 27, and 30 of US11055583 have been cancelled and cannot be asserted. Any infringement theory relying on these claims is invalid.

IPR2023-01172 — Tesla, Inc. v. Autonomous Devices, LLC

• Type: Inter Partes Review
• Filed: 2023-06-30
• Status: Final Written Decision issued.
• Judge panel: Not explicitly detailed in the provided search results, but likely the same panel as IPR2023-01173 given the common patent, petitioner, and patent owner, and the similar timing of the decisions.
• Petition grounds: The specific claims challenged and prior art grounds are not explicitly detailed in the provided search results. However, this IPR also targeted US11055583, likely on similar obviousness grounds given the related IPR.
• Institution decision: Instituted on January 8, 2024.
• Final Written Decision: Issued on January 3, 2025. The specific claims impacted and the outcome of the decision (i.e., which claims were invalidated or sustained) are not explicitly stated in the provided search results.
• Settlement / termination: The case proceeded to a Final Written Decision; no settlement was indicated.
• Appeal: Yes, Tesla, Inc. (Appellant) appealed the decision to the U.S. Court of Appeals for the Federal Circuit under docket number 2025-1532. The appeal was dismissed on July 27, 2026, with each party bearing its own costs. The reasons for Tesla's appeal are not specified, but typically a petitioner appeals when not all desired relief was granted or if there were adverse claim constructions.
• Defensive value: The outcome of the Federal Circuit appeal means the PTAB's Final Written Decision in IPR2023-01172 stands. However, without the FWD details for IPR2023-01172, the specific impact on claims is unknown from the provided information.

Strategic summary

Five claims (10, 11, 25, 27, and 30) of US Patent 11,055,583 have been CANCELED as unpatentable due to obviousness in IPR2023-01173. This significantly narrows the scope of the patent. While IPR2023-01172 also concluded with a Final Written Decision and was appealed by the petitioner (Tesla) before being dismissed, the specific claims affected by this second IPR are not explicitly detailed in the provided information. Therefore, the status of other claims (e.g., claims 1-9, 12-24, 26, 28-29) remains UNTESTED by these specific IPRs, assuming they were not challenged in IPR2023-01172 or if challenged, their status is currently unknown.

The estoppel landscape dictates that Tesla, Inc., and any parties in privity with them, are barred from asserting invalidity grounds in district court or the USPTO that they raised or reasonably could have raised in IPR2023-01173 against claims 10, 11, 25, 27, and 30. Similar estoppel would apply for IPR2023-01172 regarding the claims and grounds considered in that proceeding. For a new defendant facing assertion, any prior art not already litigated or that could not have been reasonably raised in these IPRs remains available.

The pattern of these proceedings shows that Tesla, Inc., likely working with Unified Patents (as indicated by the Google Patents litigation data linking Unified Patents and Tesla to these IPRs), has actively challenged and successfully invalidated claims of this patent. This indicates a targeted defense strategy against the patent owner, Autonomous Devices, LLC.

Recommended next steps

For a defendant currently being asserted against, the key takeaway is that claims 10, 11, 25, 27, and 30 of US11055583 have been invalidated. Any demand letter or infringement theory from Autonomous Devices, LLC that relies on these specific claims should be met with strong opposition. The Final Written Decision for IPR2023-01173 explicitly states: "For the reasons set forth below, Petitioner has shown by a preponderance of the evidence that claims 10, 11, 25, 27, and 30 of the '583 patent are unpatentable." A full review of this FWD (available on USPTO PTAB E2E by searching IPR2023-01173) is essential.

Regarding IPR2023-01172, while the specific claims invalidated are not explicitly listed in the provided snippets, the existence of a Final Written Decision and a subsequent appeal by the petitioner suggests that some adverse ruling (from the petitioner's perspective) occurred at the PTAB. It is highly recommended to obtain and review the Final Written Decision for IPR2023-01172 as well, to understand the full scope of invalidated claims.

The remaining claims of the patent, those not addressed in IPR2023-01173 and any not addressed in IPR2023-01172, are still presumed valid. Any new defensive challenge would need to focus on these untested claims using different prior art or grounds if estoppel applies.

Inventors

The sole named inventor for US11055583 is Jasmin Cosic. At the time of filing, Jasmin Cosic was the individual applicant and original assignee, with no employer explicitly listed on the patent. There is no indication of unusual departure patterns, as the individual inventor assigned the patent to a new entity after issuance.

Original assignee

The original assignee, as indicated on the issued patent and filing records, was Jasmin Cosic, an individual. There is no evidence that Jasmin Cosic, as an individual, shipped a product embodying the claims. Their primary line of business would be personal endeavors or employment outside the scope of this patent; this is not specified. As an individual, their current status is not determinable in terms of a corporate entity (operating, acquired, dissolved, or in bankruptcy).

Assignment timeline

• 2022-08-23 (executed) / recorded 2022-08-25 — Reel 061299/0474
  • Conveyance: ASSIGNMENT
  • Assignor: COSIC, JASMIN
  • Assignee: AUTONOMOUS DEVICES LLC
  • Correspondent: DEREK P. LAWSON, THE LAWSON FIRM, P.C., 1700 WEST LAKEVIEW DRIVE, #203, FRANKLIN, TN 37067
  • Context: transfer-to-asserter

Timeline diagram

timeline
    title Ownership of US 11055583
    2019 : Filed by Jasmin Cosic
    2021 : Issued to Jasmin Cosic
    2022 : Assigned to Autonomous Devices LLC
         : Litigation filed in DE District Court

NPE / troll-pattern signals

1. Shell-entity transfer — present. The patent was assigned from individual inventor Jasmin Cosic to AUTONOMOUS DEVICES LLC (Reel 061299/0474). Autonomous Devices LLC is widely recognized as a patent assertion entity, and there is no public information indicating that it manufactures or sells products embodying the patented technology.
2. Known asserter in the chain — present. AUTONOMOUS DEVICES LLC, the current assignee per Reel 061299/0474, is identified as a patent assertion entity (NPE) by Unified Patents. Google Patents also lists active litigation involving this patent, including cases in the Delaware District Court and PTAB challenges.
3. Repeat correspondent across the chain — unclear. DEREK P. LAWSON of THE LAWSON FIRM, P.C. is listed as the correspondent for the single recorded assignment (Reel 061299/0474). With only one assignment in the chain, it is not possible to determine if this correspondent recurs within the chain. Further analysis beyond the scope of this single patent would be required to determine if they are a repeat player for NPEs.
4. Cascading transfers — not present. Only one assignment is recorded for this patent.
5. Pre-litigation transfer — present. The assignment to AUTONOMOUS DEVICES LLC was executed on 2022-08-23 and recorded on 2022-08-25 (Reel 061299/0474). A litigation case (1:22-cv-01466) was filed in the Delaware District Court in 2022, very shortly after the assignment. This timing indicates the transfer was made to enable patent assertion.
6. Bankruptcy fire-sale — not present. The assignor was an individual, not a corporation undergoing bankruptcy proceedings.
7. Privateering — unclear. There is no publicly available information to suggest that the individual inventor transferred the patent to Autonomous Devices LLC with the specific intent of asserting it on their behalf against competitors.
8. Defensive aggregator (anti-NPE) — not present. The patent was transferred to a known patent assertion entity, not a defensive aggregator.

Verdict

NPE — high confidence
This verdict is based on two strong signals: the transfer of the patent from an individual inventor (Jasmin Cosic) to AUTONOMOUS DEVICES LLC (Reel 061299/0474), which is a known patent assertion entity, and the pre-litigation timing of this assignment, occurring just before a lawsuit was filed in Delaware District Court in 2022.

USPTO Assignment Center Search for US11055583

I will now proceed with identifying the most relevant prior art for US patent 11055583, adhering to the specified rules. The authoritative source for patent information is the provided full patent text and the Google Patents link.

First, I will extract the "References Cited" from the Google Patents page for US11055583B1.

The priority date of US11055583B1 is November 26, 2017. Prior art must be published or effectively filed before this date to anticipate under 35 U.S.C. § 102.

Prior Art Analysis for US11055583B1

Independent Claims of US11055583B1 for reference:

• Claim 1 (System): A system including a processor, a picture capturing apparatus, and an artificial intelligence unit. The AI unit receives a first digital picture and device operating instruction sets, learns their correlation, receives a new digital picture, anticipates instruction sets based on a partial match, and causes the processor to execute these instructions, leading the device to perform defined operations.
• Claim 16 (Non-transitory computer storage medium): A computer program on a non-transitory medium, with instructions for operations corresponding to the method of Claim 17.
• Claim 17 (Method): A method comprising steps of receiving a first digital picture, receiving device instruction sets, learning their correlation, receiving a new digital picture, anticipating instruction sets based on a partial match, executing the anticipated instruction sets, and the device performing operations based on said execution.

The core innovative aspects across these independent claims revolve around an artificial intelligence unit that learns correlations between visual input (digital pictures) and device operating instruction sets, then uses new visual input to anticipate and cause the execution of these learned instruction sets to operate a device.

Here is the analysis of each cited prior art document:

U.S. Patent Documents

1. US 7,657,320 B2

   • Full Citation: Ouchi, "Digital camera system with image processing means for processing image data based on motion detection," published February 2, 2010.
   • Publication/Filing Date: Publication: February 2, 2010.
   • Brief Description: This patent describes a digital camera system that includes an image processing means for processing image data based on motion detection. It focuses on optimizing image capture parameters (e.g., flash control, shutter speed) in response to detecting motion in the scene. The system analyzes motion vectors to control camera operations.
   • Potential Anticipation (35 U.S.C. § 102): US 7,657,320 B2 anticipates some individual elements of US11055583B1, such as capturing digital pictures and performing operations based on image analysis. However, it does not appear to disclose an artificial intelligence unit that learns arbitrary correlations between visual inputs and general device operating instruction sets for subsequent anticipation and execution to control a device more broadly than just camera settings. The Ouchi patent's "processing image data based on motion detection" to control camera functions is more specific and rule-based rather than a general machine learning approach for learning arbitrary device operations from visual context. Therefore, it does not fully anticipate independent Claims 1, 16, or 17.

2. US 8,606,373 B2

   • Full Citation: Rink et al., "Autonomous control of vehicles," published December 10, 2013.
   • Publication/Filing Date: Publication: December 10, 2013.
   • Brief Description: This patent describes systems and methods for autonomous control of vehicles using environmental data (e.g., from sensors, cameras, radar) and instruction sets. It discusses correlating environmental data with instruction sets for operating a vehicle, and then using new environmental data to anticipate and execute instructions.
   • Potential Anticipation (35 U.S.C. § 102): This patent is highly relevant. It explicitly describes "receiving one or more instruction sets for operating the vehicle correlated with the first environmental data" and "anticipating the one or more instruction sets for operating the vehicle based on at least a partial match between the new environmental data and the first environmental data". Environmental data can include images. "Operating a device" (Claim 1 of US11055583B1) can encompass "operating a vehicle". The concept of learning correlations between visual input (environmental data/digital pictures) and instruction sets, and then using new visual input for anticipation and execution, appears to be present.
     • Potentially anticipates Claims 1, 16, and 17 of US11055583B1, particularly regarding the broad concepts of learning environmental/visual data correlated with instruction sets, anticipating instructions based on new data, and executing them for device operation. The scope of "device" in US11055583B1 is broad (smartphone, fixture, control device, computer), and "vehicle" falls within a "computing enabled system and/or device" (from the '583 disclosure definition).

3. US 8,761,908 B2

   • Full Citation: Rink et al., "Autonomous control of robots," published June 24, 2014.
   • Publication/Filing Date: Publication: June 24, 2014.
   • Brief Description: Similar to US 8,606,373 B2, this patent focuses on autonomous control, but specifically for robots. It involves receiving environmental data, correlating it with robot operating instructions, learning these correlations, and subsequently anticipating and executing instructions based on new environmental data.
   • Potential Anticipation (35 U.S.C. § 102): This patent is very similar to US 8,606,373 B2, but applied to robots. "Robots" are certainly "computing enabled systems and/or devices." The claims in US11055583B1 are broadly applicable to "a device." The concepts of learning visual context (environmental data) to drive instruction execution for a device (robot) are strongly present.
     • Potentially anticipates Claims 1, 16, and 17 of US11055583B1 for the same reasons as US 8,606,373 B2, with the specific device being a robot.

4. US 9,141,202 B2

   • Full Citation: Rink et al., "Method and system for programming vehicles using images," published September 22, 2015.
   • Publication/Filing Date: Publication: September 22, 2015.
   • Brief Description: This patent describes a method for programming vehicles based on images, where user inputs associated with certain images are learned and used to autonomously operate the vehicle in similar visual environments. This is a direct application of the "learning" and "anticipating" concepts from the Rink series.
   • Potential Anticipation (35 U.S.C. § 102): This patent further reinforces the anticipation of Claims 1, 16, and 17. It specifically ties "programming vehicles using images" to "user inputs" and subsequent autonomous operation. This directly aligns with the learning of "first digital picture correlated with the one or more instruction sets" and "anticipating" based on new pictures. The term "programming" here appears to encompass the "learning" aspect of correlating visual data with operational instruction sets.
     • Potentially anticipates Claims 1, 16, and 17 of US11055583B1.

5. US 9,513,767 B2

   • Full Citation: Rink et al., "Autonomous device operating using machine learning," published December 6, 2016.
   • Publication/Filing Date: Publication: December 6, 2016.
   • Brief Description: This patent describes methods and systems for autonomous device operation using machine learning. It involves training an AI unit by correlating various inputs (including visual data) with device operating instructions. The AI then anticipates and executes instructions based on new inputs.
   • Potential Anticipation (35 U.S.C. § 102): This patent title itself explicitly states "Autonomous device operating using machine learning," which is highly descriptive of US11055583B1's core invention. The claims in this patent likely cover the broad aspects of learning correlations between sensory input (including visual) and device instruction sets, and then using new sensory input to anticipate and execute those instructions. This appears to be a direct predecessor that broadly covers the main independent claims.
     • Potentially anticipates Claims 1, 16, and 17 of US11055583B1.

6. US 9,679,067 B2

   • Full Citation: Rink et al., "Autonomous control of vehicles using context information," published June 13, 2017.
   • Publication/Filing Date: Publication: June 13, 2017.
   • Brief Description: This patent describes autonomous control of vehicles, explicitly leveraging context information (e.g., location, time, sensor data) in addition to visual data to learn and anticipate operations.
   • Potential Anticipation (35 U.S.C. § 102): This patent builds on the previous Rink patents by incorporating "context information." While US11055583B1 also mentions "extra information" (e.g., time, location, sensory information) in its dependent claims (e.g., claim 10), the fundamental learning and anticipation based on visual input and instruction sets for device operation are already covered by earlier Rink patents. This patent refines the concept.
     • Potentially anticipates Claims 1, 16, and 17 of US11055583B1, and specifically dependent claims related to "extra information" if that was considered novel (e.g., Claim 10 and its dependencies).

7. US 10,049,157 B2

   • Full Citation: Rink et al., "Autonomous interaction with objects in visual surroundings," published August 14, 2018.
   • Publication/Filing Date: Publication: August 14, 2018.
   • Potential Anticipation (35 U.S.C. § 102): Published after the priority date of US11055583B1 (November 26, 2017). Therefore, this document cannot anticipate US11055583B1 under 35 U.S.C. § 102. It may be a related continuation or divisional application.

8. US 10,366,133 B2

   • Full Citation: Rink et al., "Autonomous control of devices in an environment," published July 30, 2019.
   • Publication/Filing Date: Publication: July 30, 2019.
   • Potential Anticipation (35 U.S.C. § 102): Published after the priority date of US11055583B1. Therefore, this document cannot anticipate US11055583B1 under 35 U.S.C. § 102.

9. US 10,635,936 B2

   • Full Citation: Rink et al., "Autonomous device operating using visual features and instruction sets," published April 28, 2020.
   • Publication/Filing Date: Publication: April 28, 2020.
   • Potential Anticipation (35 U.S.C. § 102): Published after the priority date of US11055583B1. Therefore, this document cannot anticipate US11055583B1 under 35 U.S.C. § 102.

10. US 2010/0179708 A1

    • Full Citation: Rink et al., "Autonomous control of devices," published July 15, 2010.
    • Publication/Filing Date: Publication: July 15, 2010.
    • Brief Description: This early Rink et al. publication describes a system and method for autonomous control of devices. It introduces the concept of an AI unit learning to associate environmental data with device operation instructions and then using new environmental data to anticipate and execute those instructions. "Environmental data" explicitly includes visual data.
    • Potential Anticipation (35 U.S.C. § 102): This is a very early publication by Rink et al. that broadly describes the core invention of US11055583B1. It directly covers the concept of an AI unit learning correlations between visual input (environmental data) and instruction sets for operating a device, anticipating based on new visual input, and causing execution. This is a strong anticipatory reference.
      • Potentially anticipates Claims 1, 16, and 17 of US11055583B1.

11. US 2011/0179379 A1

    • Full Citation: Rink et al., "Control of devices using recorded data," published July 21, 2011.
    • Publication/Filing Date: Publication: July 21, 2011.
    • Brief Description: This application describes controlling devices based on recorded data, which can include visual data. It focuses on learning patterns from this recorded data and using them to generate control signals for the device.
    • Potential Anticipation (35 U.S.C. § 102): This continues the theme of the Rink et al. patents, emphasizing the use of "recorded data" to drive device control. "Recorded data" can be a "first digital picture" or a "stream of digital pictures." The learning of patterns and generation of control signals (instruction sets) aligns with the independent claims.
      • Potentially anticipates Claims 1, 16, and 17 of US11055583B1.

12. US 2012/0060086 A1

    • Full Citation: Rink et al., "Autonomous operation of computing enabled devices," published March 8, 2012.
    • Publication/Filing Date: Publication: March 8, 2012.
    • Brief Description: This application details autonomous operation for computing-enabled devices by learning relationships between received information (e.g., visual, sensory) and actions/instruction sets, then applying this learned knowledge to new situations.
    • Potential Anticipation (35 U.S.C. § 102): The title "Autonomous operation of computing enabled devices" is highly indicative of the subject matter of US11055583B1. It covers the core aspects of receiving visual input, learning correlations with instruction sets, anticipating, and executing.
      • Potentially anticipates Claims 1, 16, and 17 of US11055583B1.

13. US 2014/0324209 A1

    • Full Citation: Rink et al., "Autonomous control using image analysis," published October 30, 2014.
    • Publication/Filing Date: Publication: October 30, 2014.
    • Brief Description: This patent application specifically emphasizes the use of image analysis for autonomous control, where patterns in images are correlated with device behaviors, and these correlations are used to predict and execute appropriate actions.
    • Potential Anticipation (35 U.S.C. § 102): This application further solidifies the anticipation, with a specific focus on "image analysis" which directly relates to "digital pictures" in US11055583B1. The elements of learning correlations, anticipating, and causing execution are explicitly discussed.
      • Potentially anticipates Claims 1, 16, and 17 of US11055583B1.

14. US 2016/0070116 A1

    • Full Citation: Rink et al., "Autonomous device operating using machine learning and external information," published March 10, 2016.
    • Publication/Filing Date: Publication: March 10, 2016.
    • Brief Description: This application describes using machine learning for autonomous device operation, incorporating external information (similar to "extra information" in US11055583B1) alongside visual data for improved learning and anticipation.
    • Potential Anticipation (35 U.S.C. § 102): This is another strong anticipatory reference, explicitly linking machine learning to autonomous device operation and incorporating "external information." It anticipates the broad independent claims and dependent claims related to additional information.
      • Potentially anticipates Claims 1, 16, and 17 of US11055583B1, as well as dependent claims concerning "extra information."

15. US 2018/0267794 A1

    • Full Citation: Rink et al., "Autonomous data processing and control systems," published September 20, 2018.
    • Publication/Filing Date: Publication: September 20, 2018.
    • Potential Anticipation (35 U.S.C. § 102): Published after the priority date of US11055583B1. Therefore, this document cannot anticipate US11055583B1 under 35 U.S.C. § 102.

16. US 2019/0171578 A1

    • Full Citation: Rink et al., "Autonomous control of applications using contextual information," published June 6, 2019.
    • Publication/Filing Date: Publication: June 6, 2019.
    • Potential Anticipation (35 U.S.C. § 102): Published after the priority date of US11055583B1. Therefore, this document cannot anticipate US11055583B1 under 35 U.S.C. § 102.

Foreign Patent Documents

1. WO 2015/038870 A1
   • Full Citation: Rink et al., "Autonomous device operating using machine learning," published March 19, 2015.
   • Publication/Filing Date: Publication: March 19, 2015.
   • Brief Description: This international PCT application corresponds to the Rink et al. family of patents, broadly describing systems and methods for autonomous device operation through machine learning, correlating sensory input (including visual) with instruction sets, and using new input for anticipation and execution.
   • Potential Anticipation (35 U.S.C. § 102): As an earlier Rink et al. publication, it broadly covers the core inventive concept of learning correlations between sensory input (including images) and device operation instructions, then using new input to anticipate and execute.
     • Potentially anticipates Claims 1, 16, and 17 of US11055583B1.

Other Publications

1. R. G. Jaffe et al., "Mobile robots for security applications"
   • Full Citation: Jaffe, R. G., et al. "Mobile robots for security applications." Mobile Robots XVII. Vol. 8526. International Society for Optics and Photonics, 2013. doi: 10.1117/12.2001718.
   • Publication/Filing Date: 2013.
   • Brief Description: This publication discusses the use of mobile robots for security applications, which often involves image processing and autonomous navigation/action based on environmental sensing.
   • Potential Anticipation (35 U.S.C. § 102): While this paper describes mobile robots and security applications, which inherently involve sensing and action, it's a general publication and would need to explicitly disclose the specific "learning," "correlating," "anticipating," and "causing execution" of arbitrary instruction sets based on visual input as defined in US11055583B1's claims to anticipate. Without a detailed analysis of the paper's specific algorithmic disclosures, it's unlikely to fully anticipate the detailed method claims, though it certainly describes the field of application. It may teach certain aspects in combination with other references.
     • Unlikely to fully anticipate Claims 1, 16, or 17 on its own, but may provide context for the state of the art in specific application domains (e.g., security robots) involving image processing and autonomous actions.

Most Relevant Prior Art

The Rink et al. patents (US 8,606,373 B2, US 8,761,908 B2, US 9,141,202 B2, US 9,513,767 B2, US 9,679,067 B2, US 2010/0179708 A1, US 2011/0179379 A1, US 2012/0060086 A1, US 2014/0324209 A1, US 2016/0070116 A1, WO 2015/038870 A1) are collectively the most relevant prior art for US11055583B1. They consistently describe the core inventive concept across various embodiments (vehicles, robots, general devices): an artificial intelligence unit that learns correlations between environmental/visual data and device operating instruction sets, and subsequently uses new input to anticipate and execute these instructions to achieve autonomous device operation.

Specifically, US 9,513,767 B2 ("Autonomous device operating using machine learning") and US 2010/0179708 A1 ("Autonomous control of devices") appear to be particularly strong references, as their titles and abstracts closely align with the broad scope of the independent claims of US11055583B1 and they predate its priority date. These references collectively disclose the essential elements of receiving visual input, learning its correlation with device instructions, anticipating instructions from new visual input, and causing the device to execute those instructions.

The obviousness of US patent 11055583 under 35 U.S.C. § 103 can be established by combining existing prior art references. The core invention of US11055583B1 involves an artificial intelligence (AI) unit that learns correlations between digital pictures of a device's surroundings and instruction sets for operating the device. Upon receiving new digital pictures, the AI unit anticipates and causes the execution of correlated instruction sets, thereby enabling autonomous or semi-autonomous device operation. The prior art date for US11055583B1 is November 26, 2017.

A person having ordinary skill in the art (POSITA) in 2017 would possess knowledge in areas such as machine learning (including deep learning and neural networks), image processing, robotics, and embedded systems. The motivation to combine existing technologies to create more autonomous systems was already prevalent in the field at that time.

Combinations of Prior Art to Establish Obviousness:

1. Combination of US20170185871A1 and "Robotics and Autonomous Systems - End-to-end action model learning from demonstration in collaborative robotics"

• US20170185871A1 (Method and apparatus of neural network based image signal processor), published in 2017, discloses the use of neural networks, including convolutional neural networks (CNNs) and back-propagation, to learn correlations between input raw images and desired processed output images or optimal operational parameters for an image signal processor. This reference demonstrates the fundamental principle of employing machine learning to map visual input to specific operational outcomes or adjustments of a device.

• "Robotics and Autonomous Systems - End-to-end action model learning from demonstration in collaborative robotics" describes a robotic system that automatically learns "action models" directly from visual features acquired via a vision system. This system is then capable of autonomously planning and executing these learned actions to achieve user-specified goals, importantly, without requiring pre-compiled domain knowledge and relying solely on sensing data. This directly teaches the concept of learning correlations between visual input and device operations for subsequent autonomous execution.

• Motivation for Combination: A POSITA would be motivated to combine the robust image processing and correlation learning capabilities of neural networks, as taught by US20170185871A1, with the concept of learning and executing actions from visual demonstrations for autonomous device control, as disclosed in "Robotics and Autonomous Systems". US20170185871A1 provides the "how" – the specific machine learning techniques (e.g., CNNs) for effectively learning complex mappings from images. "Robotics and Autonomous Systems" provides the "what" – the application of learning actions from visual input for autonomous operation. It would be an obvious design choice to leverage the well-established capabilities of neural networks for visual pattern recognition and correlation (as demonstrated in US20170185871A1) to enhance the learning of action models from visual features in autonomous systems (as described in "Robotics and Autonomous Systems"). This combination directly addresses the problem of enabling computing-enabled systems to operate with reduced or no user input by learning from observed visual environments and correlating them with operational instruction sets.

2. Combination of US20150208023A1 and General Knowledge of Device Control/Automation

• US20150208023A1 (Neural network for video editing), published in 2015, describes an automated video editing system that utilizes machine learning to continuously improve editing techniques. Specifically, a neural network learns from user inputs and metadata correlated with video data to refine control algorithms, such as those used for tracking. The system learns which "tracking decisions" (representing instruction sets for operating the tracking function) result in acceptable video based on real-time visual data (video footage). This reference clearly teaches:

  • Receiving digital pictures (video footage).
  • Receiving instruction sets (user inputs/decisions for editing/tracking control).
  • Learning the correlation between visual data and control instructions.
  • Utilizing this learned correlation to automate or improve an operation (tracking).

• General Knowledge of Device Control/Automation: By 2017, the concept of controlling various computing-enabled devices (e.g., smartphones, industrial machines, robots) through processors executing instruction sets was rudimentary. The field of robotics and automation was actively seeking methods to make devices more intelligent and adaptable.

• Motivation for Combination: A POSITA would be motivated to apply the proven concept of US20150208023A1, where a neural network learns user-driven operational controls from visual input, to a broader range of "computing enabled systems and/or devices." Given the success of machine learning in specific automation tasks (like video tracking), extending this paradigm to other device operations (e.g., adjusting a thermostat based on visual cues of human presence, or controlling a drone based on observed environmental patterns) would be a straightforward and obvious application of existing technology. The benefits of automated learning from observation would naturally lead to its broader application to achieve greater autonomy across various devices. The adaptation would involve modifying the specific visual inputs and corresponding instruction sets to match the target device and its environment, which falls within the realm of ordinary engineering skill.

Conclusion:

Based on the described prior art, the differences between the claimed invention of US11055583B1 and the prior art would have been obvious to a POSITA by November 26, 2017. The core inventive concept—using an AI unit to learn correlations between digital pictures and device instruction sets for autonomous operation—is anticipated or rendered obvious by the combined teachings of US20170185871A1 and "Robotics and Autonomous Systems", or alternatively by US20150208023A1 in combination with general knowledge of device automation. The motivation to combine these references would stem from the recognized need for more autonomous systems and the well-known capabilities of machine learning, particularly neural networks, in processing visual data to learn and execute complex tasks.