PTAB

IPR2013-00095

Oracle Corp v. Clouding IP LLC

Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: System and Method for Automatically Maintaining a Computer System
  • Brief Description: The ’839 patent describes a maintenance tool that uses a set of software sensors and an artificial intelligence (AI) engine with a case-base knowledge database to automatically diagnose and resolve computer system problems.

3. Grounds for Unpatentability

Ground 1: Obviousness over Gurer - Claims 1, 2, 6, 8, 14, 15, and 17 are obvious under 35 U.S.C. §103 over Gurer.

  • Prior Art Relied Upon: Gurer (An Artificial Intelligence Approach to Network Fault Management, published Feb. 1996).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Gurer, which was not considered during prosecution, teaches all elements of the challenged claims. Gurer discloses an automated network fault management system using case-based reasoning (CBR) and AI. This system uses "alarms" (argued to be equivalent to the claimed "sensors") from network elements to detect problems. It employs a "case library" (the claimed "knowledge database") with prior problems and solutions. The system interprets data from the alarms against the case library to diagnose faults. Crucially, Gurer teaches an iterative process where if more data is needed, the system sends tests to network elements (i.e., "activates a particular sensor") to gather more information. Gurer also explicitly teaches that when a fault is identified, the entire problem-solving episode, including tests and results, is stored as a "new case" in the case library, directly mapping to the key limitation of saving data for a new case when a solution is not found.
    • Key Aspects: Petitioner contended Gurer's system, focused on network fault management, is directly analogous to the ’839 patent’s general computer system maintenance, and a person of ordinary skill in the art (POSITA) would have readily applied Gurer's teachings to the broader context.

Ground 2: Obviousness over Allen ’218 - Claims 1, 2, 6, 8, 14, 15, and 17 are obvious over Allen ’218.

  • Prior Art Relied Upon: Allen ’218 (Patent 5,586,218).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner asserted that Allen ’218, also not previously considered, describes a "software agent" (AI engine) that performs autonomous learning in an environment (a computer system) using a CBR system. The agent is coupled to a "sensor" for gathering information and an "effector" for manipulating the environment. The system includes a "case base" (knowledge database). When the system needs more information, a "behavior module" generates a "queries message" to request further information, which Petitioner argued meets the limitation of activating a sensor to gather data. Allen ’218 further discloses that "new cases may be produced by inspection of scenarios from the environment," satisfying the limitation of saving data as a new case.
    • Motivation to Combine (for Modification): Allen ’218 discloses a single "sensor." Petitioner argued that modifying this to a "plurality of sensors" as claimed would have been an obvious design choice for a POSITA. Using multiple sensors to gather more comprehensive data from different aspects of a computer system was a well-known and predictable way to improve the performance and accuracy of any diagnostic system.
    • Expectation of Success: A POSITA would have had a high expectation of success in using multiple sensors, as it was a simple, common-sense approach to gathering more complete information for a diagnostic AI.

Ground 3: Obviousness over Barnett in view of Allen ’664 - Claims 6, 8, and 14 are obvious over Barnett in view of Allen ’664.

  • Prior Art Relied Upon: Barnett (Patent 5,664,093) and Allen ’664 (Patent 5,581,664).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Barnett discloses a fault management system for a distributed computer system that meets most limitations of the method claims. Barnett’s system uses a plurality of "measurement agents" (sensors) to obtain performance information, a "diagnostic system" (AI engine) that is activated upon fault detection, and a "rule base" to identify faults and provide solutions. However, Petitioner contended Barnett is silent on what to do when its rule-based system cannot determine a likely solution. Allen ’664, which teaches a hybrid CBR/rule-based system, was argued to supply this missing element. Allen ’664 explicitly discloses that when its inference engine cannot find a matching case (solution), it can "invoke a rule" or "create a new case," thereby teaching the step of saving a state of the computer system when a solution is not found.
    • Motivation to Combine: A POSITA would combine these references to improve Barnett's system. Barnett's purely rule-based system is limited to known problems. A POSITA would have been motivated to upgrade Barnett’s rule-based engine with the hybrid CBR/rule-based approach from Allen ’664 to gain the well-known advantages of CBR, such as the ability to learn from new situations and handle problems not covered by existing rules.
    • Expectation of Success: A POSITA would have had a reasonable expectation of success because Allen ’664 demonstrates how CBR can be smoothly integrated with a rule-based system. The combination represented a known technique for improving a known system to achieve a predictable result (enhanced learning and adaptability).

4. Key Claim Construction Positions

  • Petitioner argued for a broad construction of key terms under the Broadest Reasonable Construction standard.
  • "Sensors": This term should be interpreted to include different aspects of the same software program or different components of the same application that gather information.
  • "Artificial Intelligence (AI) Engine": This term should be interpreted as a different aspect of the same software program as the "sensors."
  • These constructions were central to Petitioner's ability to map the functionality described in the prior art—where fault management might be a single, integrated software application—to the distinct "sensor" and "AI engine" limitations of the claims.

5. Relief Requested

  • Petitioner requested the institution of an inter partes review and cancellation of claims 1, 2, 6, 8, 14, 15, and 17 of the ’839 patent as unpatentable.