US 5704012

US Patent 5704012, titled "Adaptive resource allocation using neural networks," was issued to International Business Machines Corp on December 30, 1997, based on a filing date of May 31, 1995. The sole inventor listed is Joseph Phillip Bigus.

Abstract:
The patent describes a resource allocation controller that is customized to a system's available resources and configuration. This controller dynamically allocates resources or alters the configuration to manage changing workloads. In its preferred embodiment, this controller is part of a computer's operating system and utilizes a controller neural network for decision-making and a separate system model neural network for modeling the system and training the controller. The system collects performance data to train the system model. A system administrator sets performance targets, and any deviations from these targets are fed back through the system model to the controller neural network, creating a closed-loop system for adaptive resource allocation.

Plain-Language Overview of Independent Claims:

Claim 1:
This claim outlines a method for controlling how a computer system responds to its workload and configuration. The method involves several steps:

1. Gathering Performance Data: The system collects information on how different categories of jobs (which require varying computer resources) perform over time. This data includes details about the workload, the system's configuration, and the computer system's actual response.
2. Constructing a Neural Network: A neural network is built with inputs that represent the computer system's workload and configuration, and at least one output that represents the system's response.
3. Training the Neural Network: This neural network is trained using the collected performance data to create an accurate model of the computer system's behavior.
4. Determining System Response: The trained neural network is then used to predict the computer system's response.
5. Allocating Resources: Based on these predicted responses and performance goals specified by a user for each job category, the computer system's resources are allocated among the different job classes.

Legal Status and CAFC Docket Search:
According to the patent information, US patent 5704012 is "Expired - Fee Related" with an anticipated expiration date of December 30, 2014. The patent officially "Lapsed for failure to pay maintenance fees" with an effective date of January 30, 2002.
A search of the CAFC 2026 dockets, specifically the "Scheduled Cases – June 2026," did not reveal any cases mentioning patent number 5704012. Given the patent's expired status since 2002 due to maintenance fee non-payment, it is highly unlikely to be involved in active litigation.

A search for litigation involving US patent 5704012 across patent litigation databases like Unified Patents, CAFC, and PACER has not yielded any known cases.

Therefore, based on the current search results, there is no known litigation involving US patent 5704012.

Proceedings overview

There are no AIA trial proceedings on file for US Patent 5704012. This means the patent has not been subjected to Inter Partes Review (IPR), Post-Grant Review (PGR), or Covered Business Method (CBM) patent review proceedings at the Patent Trial and Appeal Board (PTAB). This gives a defendant no specific PTAB-based defensive posture related to invalidated claims, as no claims have been challenged or canceled through these proceedings.

Strategic summary

As of the current date, no claims of US Patent 5704012 have been canceled or sustained through PTAB proceedings. All claims remain untested by the PTAB. There is no estoppel landscape established by PTAB decisions for this patent. The absence of PTAB activity suggests that the patent has either not been aggressively asserted in a manner that would provoke IPRs, or potential petitioners have not found viable grounds for challenge.

Recommended next steps

Since no PTAB activity exists for US Patent 5704012, a defendant facing assertion of this patent would need to:

• Conduct a thorough prior art search to identify potential grounds for an IPR or PGR if the patent's claims are relevant to their activities.
• Evaluate the strength of such grounds against the patent's single claim to determine the feasibility of filing a PTAB petition.
• Consider that the absence of PTAB challenges for a patent granted in 1997 (and expired in 2002) might indicate either a lack of commercial relevance or a lack of strong prior art grounds.

Inventors

The sole inventor listed for US Patent 5704012 is Joseph Phillip Bigus. He was an employee of International Business Machines Corp. at the time of filing, as indicated by the original assignee. No unusual patterns, such as the inventor departing the original assignee around the filing date, are determinable from the provided patent text.

Original assignee

The entity named on the issued patent is International Business Machines Corp.

IBM is a multinational technology and consulting company. At the time of the patent's filing and issuance, IBM was a major manufacturer and vendor of computer hardware (e.g., the IBM Application System/400 midrange computer mentioned in the patent), software (including operating systems), and services. The patent describes an adaptive resource allocation system using neural networks as part of a computer's operating system, indicating that IBM likely shipped products embodying the claims as part of its computer systems and operating system offerings.

IBM's current status is operating.

Assignment timeline

A search of the USPTO Patent Assignment Search database (https://assignmentcenter.uspto.gov/) for US Patent 5704012 yielded no recorded assignments. This indicates that, according to USPTO records, the patent has remained with its original assignee, International Business Machines Corp., since its issuance.

Timeline diagram

timeline
    title Ownership of US 5704012
    1995 : Filed by IBM
    1997 : Issued to IBM
    2002 : Lapsed for fee non-payment
    2014 : Anticipated expiration

NPE / troll-pattern signals

1. Shell-entity transfer — not present. The patent remains with International Business Machines Corp., an operating company.
2. Known asserter in the chain — not present. The patent remains with International Business Machines Corp.
3. Repeat correspondent across the chain — not present. There are no recorded assignments to create a chain.
4. Cascading transfers — not present. There are no recorded assignments.
5. Pre-litigation transfer — not present. There are no recorded assignments and no known litigation.
6. Bankruptcy fire-sale — not present. International Business Machines Corp. has not undergone bankruptcy proceedings that would typically lead to such a sale.
7. Privateering — not present. The patent remains with International Business Machines Corp.
8. Defensive aggregator (anti-NPE) — not present. The patent has not been acquired by a defensive aggregator.

Verdict

Insufficient data.
There are no recorded assignments for US Patent 5704012 in the USPTO assignment database, indicating that the patent has remained with its original assignee, International Business Machines Corp., since its issuance. Therefore, there is no assignment chain to analyze for NPE/troll-pattern signals. The patent is also expired due to maintenance fee non-payment as of January 30, 2002.

USPTO Patent Assignment Search: https://assignmentcenter.uspto.gov/

Most Relevant Prior Art for US Patent 5704012

The following analysis identifies the most relevant prior art cited within US Patent 5704012, focusing on their potential to anticipate the single independent claim under 35 U.S.C. § 102. The information for each cited patent is drawn directly from the US5704012 patent document, as direct database searches for these specific older patent numbers via the search tool yielded unrelated results.

Claim 1 of US5704012 outlines a method for controlling the response of a computer system to a workload and configuration, comprising the steps of:

1. Gathering performance data (workload, configuration, response data for job classes).
2. Constructing a neural network (inputs: workload, configuration; output: system response).
3. Training the neural network with performance data to model the computer system.
4. Determining computer system response from the trained neural network's output.
5. Allocating resources among job classes based on the determined response and user-specified performance objectives.

---

1. US5325525A (Hewlett-Packard Company)

• Full Citation: US5325525A, "Method of automatically controlling the allocation of resources of a parallel processor computer system by calculating a minimum execution time of a task and scheduling subtasks against resources to execute the task in the minimum time."
• Publication/Filing Date: Publication Date: 1994-06-28; Priority Date: 1991-04-04.
• Brief Description: This patent describes a method for automatically controlling resource allocation in a parallel processor computer system by calculating a minimum execution time for a task and scheduling subtasks to achieve that time.
• Potential Anticipation (35 U.S.C. § 102): While this patent addresses resource allocation in a computer system, its mechanism for control involves calculating minimum execution times and scheduling subtasks, not the construction and training of a neural network to model system performance and then using that model for allocation based on user objectives. Thus, it does not anticipate all elements of Claim 1.

2. US5483468A (International Business Machines Corporation)

• Full Citation: US5483468A, "System and method for concurrent recording and displaying of system performance data."
• Publication/Filing Date: Publication Date: 1996-01-09; Priority Date: 1992-10-23.
• Brief Description: This patent describes a system and method focused on concurrently recording and displaying system performance data.
• Potential Anticipation (35 U.S.C. § 102): This patent discloses aspects related to "gathering performance data" (step 1 of Claim 1 of US5704012). However, it does not include the subsequent inventive steps of constructing, training, and utilizing a neural network for performance prediction and resource allocation, nor the concept of user-defined performance objectives. Therefore, it does not anticipate Claim 1.

3. US5598076A (Siemens Aktiengesellschaft)

• Full Citation: US5598076A, "Process for optimizing control parameters for a system having an actual behavior depending on the control parameters."
• Publication/Filing Date: Publication Date: 1997-01-28; Priority Date: 1991-12-09.
• Brief Description: This patent describes a process for optimizing control parameters in a system where behavior is dependent on those parameters.
• Potential Anticipation (35 U.S.C. § 102): This patent generally describes optimizing system control parameters, which broadly relates to the objective of US5704012. However, the description in US5704012 does not indicate that this process specifically involves neural networks for modeling computer system performance or for resource allocation within job classes based on user objectives. Without these specific elements, it does not anticipate Claim 1.

4. US5067107A (Hewlett-Packard Company)

• Full Citation: US5067107A, "Continuous computer performance measurement tool that reduces operating system produced performance data for logging into global, process, and workload files."
• Publication/Filing Date: Publication Date: 1991-11-19; Priority Date: 1988-08-05.
• Brief Description: This patent describes a tool for continuous computer performance measurement, data reduction, and logging into specific files (global, process, and workload).
• Potential Anticipation (35 U.S.C. § 102): Similar to US5483468A, this patent focuses on the collection and storage of computer system performance data (part of step 1 of Claim 1). It does not extend to the use of neural networks for modeling system response or for dynamically allocating resources based on user objectives. Thus, it does not anticipate Claim 1.

5. US5235673A (International Business Machines Corporation)

• Full Citation: US5235673A, "Enhanced neural network shell for application programs."
• Publication/Filing Date: Publication Date: 1993-08-10; Priority Date: 1991-04-18.
• Brief Description: US Patent 5704012 incorporates this patent by reference for describing the operation of the IBM Neural Network Utility, which simulates neural networks. This patent pertains to providing a framework or utility for implementing neural networks in application programs.
• Potential Anticipation (35 U.S.C. § 102): This patent is highly relevant as it describes the fundamental neural network technology (steps 2 and 3 of Claim 1) that US5704012 leverages. It provides the "shell" for constructing and training neural networks. However, it describes a generic utility rather than the specific application of such a neural network to model computer system performance and allocate computer resources among job classes based on user-specified performance objectives. Therefore, while foundational, it does not anticipate the complete method of Claim 1.

6. US5485545A (Mitsubishi Denki Kabushiki Kaisha)

• Full Citation: US5485545A, "Control method using neural networks and a voltage/reactive-power controller for a power system using the control method."
• Publication/Filing Date: Publication Date: 1996-01-16; Priority Date: 1991-06-20.
• Brief Description: This patent describes a control method that employs neural networks, specifically in the context of a voltage/reactive-power controller within a power system.
• Potential Anticipation (35 U.S.C. § 102): This patent is significant because it explicitly teaches the use of "neural networks" in a "control method." This demonstrates that the concept of neural network-based control was known prior to US5704012. However, the specific application is to a "power system" for electrical control, which is a different technical domain than "controlling the response of a computer system to a workload and configuration" through resource allocation among "job classes" with "performance objectives." Due to this difference in the system being controlled and the nature of the resources, it does not anticipate all elements of Claim 1.

---

Conclusion on Most Relevant Prior Art:

The most relevant prior art references, in terms of the underlying technology, are US5235673A (neural network shell) and US5485545A (neural network control in a different system). US5235673A provides the generic toolset for building neural networks, which are then specifically applied in US5704012. US5485545A demonstrates that using neural networks for control was known. However, none of the cited prior art documents appear to directly anticipate Claim 1 of US5704012 under 35 U.S.C. § 102 because they do not individually disclose all the elements of the claim, particularly the novel combination of using a neural network to model a computer system's performance and then specifically using that model to allocate computer system resources among job classes based on user-specified performance objectives.

Obviousness Analysis of US Patent 5704012 Under 35 U.S.C. § 103

This analysis considers whether the single independent claim (Claim 1) of US Patent 5704012 would have been obvious to a person having ordinary skill in the art (PHOSITA) at the time of the invention (priority date: October 8, 1993), based on the prior art identified in the "Prior Art" section.

Claim 1 of US5704012 outlines a method for controlling the response of a computer system to a workload and configuration, comprising the steps of:

1. Gathering performance data for jobs in a plurality of job classes (workload, configuration, response data for a plurality of time intervals), where jobs require different amounts of computer system resources.
2. Constructing a neural network, with inputs corresponding to the workload and configuration, and at least one output corresponding to the computer system's response.
3. Training said neural network with the gathered performance data to produce a trained neural network model of the computer system.
4. Determining the response of the computer system from the output of the trained neural network.
5. Allocating the resources in the computer system among the plurality of job classes based on the determined response and user-specified performance objectives for each job class.

Combination of Prior Art References and Rationale for Obviousness

A combination of the following prior art references would render Claim 1 of US5704012 obvious to a PHOSITA:

• US5483468A (IBM): "System and method for concurrent recording and displaying of system performance data."
• US5067107A (Hewlett-Packard Company): "Continuous computer performance measurement tool..."
• US5235673A (IBM): "Enhanced neural network shell for application programs."
• US5485545A (Mitsubishi Denki Kabushiki Kaisha): "Control method using neural networks and a voltage/reactive-power controller for a power system using the control method."
• US5325525A (Hewlett-Packard Company): "Method of automatically controlling the allocation of resources of a parallel processor computer system..."
• US5598076A (Siemens Aktiengesellschaft): "Process for optimizing control parameters for a system..."

Motivation for Combination:

The background of US5704012 itself highlights the existing problems in resource allocation, noting the difficulty of traditional heuristic and queueing theory models to adapt to constantly changing computer system configurations and workloads. It explicitly states a need for more flexible and dynamic resource allocation, and for enhanced techniques for managing system resources. A PHOSITA, faced with these known challenges, would be motivated to seek more adaptive and intelligent control solutions. Neural networks, known for their ability to learn complex, non-linear relationships and adapt to changing conditions, would be a natural choice for such an improvement.

How the Combination Renders Claim 1 Obvious:

1. Gathering performance data (Claim 1, Step 1):

   • US5483468A explicitly teaches "concurrent recording and displaying of system performance data" within a computer system. This includes collecting information regarding system performance.
   • US5067107A further details a "continuous computer performance measurement tool that reduces operating system produced performance data for logging into global, process, and workload files."
   • These references clearly establish that the gathering of comprehensive computer system performance data, including workload, configuration, and response data for various job classes, was well-known in the art. A PHOSITA would routinely collect such data to monitor and understand computer system behavior.

2. Constructing a neural network (Claim 1, Step 2):

   • US5235673A (assigned to IBM, the same assignee as US5704012) describes an "enhanced neural network shell for application programs," which provides the underlying technology for constructing neural networks. This patent teaches how to define the type and topology of a neural network, including the number of inputs, outputs, and connections.
   • US5485545A demonstrates the application of neural networks in "control method[s]" where the neural network would inherently have inputs representing system parameters to be controlled and outputs representing control actions or system responses.
   • A PHOSITA, desiring to model the complex relationships between computer system parameters (workload, configuration) and performance (response), would find it obvious to apply the general neural network construction techniques of US5235673A, designing the network with inputs corresponding to the collected computer system workload and configuration data (from US5483468A) and outputs corresponding to the system's response.

3. Training said neural network (Claim 1, Step 3):

   • US5235673A clearly teaches the training of neural networks for application programs.
   • The combination of the gathered performance data (from US5483468A) with the neural network training capability (from US5235673A) to create a model that learns the behavior of the computer system is a straightforward application of neural network technology. The motivation to create an accurate and adaptive model, as described in US5704012's background, would naturally lead a PHOSITA to train the constructed neural network with available historical performance data.

4. Determining the response of said computer system (Claim 1, Step 4):

   • Once a neural network is constructed and trained to model a system's behavior (as covered by the preceding steps), its fundamental function is to predict or "determine the response" based on new inputs.
   • US5485545A shows a neural network being used to predict system behavior (e.g., voltage/reactive-power) in a power system for control purposes.
   • Therefore, using the output of the trained neural network model of the computer system to determine its response is an inherent and obvious consequence of having such a model.

5. Allocating the resources in said computer system (Claim 1, Step 5):

   • US5325525A teaches methods for "automatically controlling the allocation of resources of a parallel processor computer system." This demonstrates that resource allocation in computer systems was a known problem and objective.
   • US5598076A describes a "process for optimizing control parameters for a system having an actual behavior depending on the control parameters." This provides a general teaching of using feedback to optimize system control.
   • US5485545A teaches using a neural network within a "control method" to effect changes in a system (a power system).
   • A PHOSITA, observing that a trained neural network can accurately predict computer system response (steps 1-4), and knowing the general art of resource allocation (US5325525A) and adaptive control using neural networks (US5485545A), would be motivated to integrate these to dynamically allocate resources. The "user-specified performance objectives" are a standard input for any control system aiming to optimize performance in a computer environment, providing the target for the resource allocation decisions. The adaptive nature of neural networks would make them a desirable component for dynamically adjusting resources to meet these objectives in the face of changing workloads, addressing the very problem articulated in the background of US5704012.

Conclusion:

Given the widespread knowledge of computer system performance monitoring and data collection (US5483468A, US5067107A), the availability of generic neural network development tools (US5235673A, from the same assignee), and the known application of neural networks for control in other complex systems (US5485545A), it would have been obvious for a PHOSITA to combine these elements. The motivation would be to develop a more adaptive and effective resource allocation mechanism for computer systems, addressing the known limitations of existing techniques, by using a neural network to model system behavior and drive resource allocation decisions based on user-defined performance objectives, thus achieving the known goal of optimized system performance. Therefore, Claim 1 of US5704012 is rendered obvious by the combination of these prior art references.

For US Patent 5704012, the following details are available from the patent document and Google Patents data:

• Patent Term Adjustments (PTA): There is no information provided in the patent text or on the Google Patents page regarding Patent Term Adjustments (PTA). PTA calculations, as they are commonly understood today, were introduced after the Uruguay Round Agreements Act (URAA) in 1995 and became effective with the American Inventors Protection Act (AIPA) in 1999. Given the patent's issuance in 1997 and its lapse in 2002, PTA would not be a significant factor.
• Patent Term Extensions (PTE): There is no information provided in the patent text or on the Google Patents page regarding Patent Term Extensions (PTE). PTE typically applies to patents covering products subject to regulatory review periods, which is not applicable to this patent's subject matter.
• Continuation Applications: The patent document states, "The present application is related to commonly assigned copending U.S. patent application Ser. No. 08/134,764, filed Oct. 8, 1993, to Bigus, entitled "Adaptive Job Scheduling Using Neural Network Priority Functions"". While this is a related application, the current patent, US5704012 (application number US08/454,977), is not explicitly identified as a continuation of it.
• Divisional Applications: US Patent 5704012 is a divisional application. The patent text explicitly states, "This is a divisional of application Ser. No. 08/134,953 filed on Oct. 8, 1993, abandoned".
• Related Family Members:
  • US08/134,953: The parent application from which US5704012 (US08/454,977) was divided. This application was filed on October 8, 1993, and is noted as abandoned.
  • US08/134,764: A commonly assigned copending U.S. patent application, filed on October 8, 1993, titled "Adaptive Job Scheduling Using Neural Network Priority Functions".
  • US5745652A (Application number US08/455,314): This patent is listed under "Family Applications" and "Also Published As" and shares the same priority date (October 8, 1993) and filing date (May 31, 1995) as US5704012.
• Projected Expiration Date:
  • The patent was published on December 30, 1997.
  • Google Patents indicates an "Anticipated expiration" date of December 30, 2014. This would typically be 20 years from the filing date (May 31, 1995), plus any adjustments, or from the priority date depending on the law at the time. However, the effective expiration date is much earlier.
  • The patent officially "Lapsed for failure to pay maintenance fees" with an effective date of January 30, 2002. Therefore, the patent is already expired.

Defensive Disclosure: Adaptive Resource Allocation Using Neural Networks (US5704012)

This defensive disclosure aims to broaden the scope of existing prior art related to adaptive resource allocation using neural networks, making future incremental improvements in this domain obvious or non-novel. The derivations explore variations in materials, operational parameters, cross-domain applications, integration with emerging technologies, and failure modes, all grounded in the principles outlined in US Patent 5704012, specifically Claim 1.

Core Claim 1 of US5704012 (Summary for reference):

A method for controlling the response of a computer system to a workload and configuration, comprising: gathering performance data; constructing a neural network (inputs: workload, configuration; output: system response); training the neural network with performance data to model the computer system; determining system response from the trained network; and allocating resources based on the determined response and user-specified performance objectives.

---

Derivative Variations

1. Material & Component Substitution: Neuromorphic Resource Orchestration with Non-Volatile Memory

Enabling Description:
A method for adaptive resource allocation in a computer system utilizing a neuromorphic processing unit for neural network execution and non-volatile memory (NVM) for dynamic resource pooling. Performance data, including instruction per cycle (IPC) rates, cache hit/miss ratios, memory access latencies, and inter-core communication overhead, is gathered by a dedicated Field-Programmable Gate Array (FPGA)-based monitoring unit. This high-resolution performance data is streamed to a neural network, emulated on a neuromorphic chip (e.g., Intel Loihi or IBM TrueNorth), which is specifically designed to model the computer system's behavior with event-driven, sparse activity patterns. The neuromorphic neural network is trained using an online, asynchronous reinforcement learning algorithm, directly adapting its synaptic weights based on observed system state transitions. The predicted system response from the neuromorphic network guides a resource manager, implemented as a custom hardware accelerator (e.g., an ASIC), to allocate resources. These resources include dynamically partitioning portions of a shared Phase-Change Memory (PCM) or Resistive RAM (ReRAM) array among various job classes, adjusting memory access priority queues, and configuring specialized interconnects to meet user-defined Service Level Objectives (SLOs) such as guaranteed throughput or maximum latency thresholds.

graph TD
    A[Workload Request Streams] --> B(FPGA Performance Monitor<br>@ System Interconnect)
    B --> C{Event-Driven Performance Data}
    C --> D[Neuromorphic Processing Unit<br>(Spiking Neural Network Model)]
    D -- Trained Model Updates --> E(ASIC Resource Manager<br>Hardware Accelerator)
    F[User SLOs<br>(Latency, Throughput)] --> E
    E --> G[NVM Resource Pools<br>(PCM/ReRAM Partitions)]
    G --> H[Computer System Core Resources<br>(CPU, GPU, Interconnect)]
    H -- Actual Performance Feedback --> B

2. Operational Parameter Expansion: Ultra-Low-Latency, Hyper-Scale Resource Allocation

Enabling Description:
A method for ultra-low-latency, hyper-scale resource allocation within a distributed computing environment, specifically an exascale supercomputing cluster comprising millions of interconnected processing elements (CPUs, GPUs, FPGAs). Performance data, including sub-microsecond-level process latencies, inter-node communication contention, fabric bandwidth utilization, and on-die temperature differentials, is gathered at a frequency exceeding 100 kHz across all computational nodes. A massively parallelized Spiking Neural Network (SNN) model, distributed across multiple GPU clusters (e.g., NVIDIA DGX systems interconnected via NVLink), is constructed and trained on this high-frequency, high-volume telemetry. The SNN's inputs include real-time distributed workload patterns (e.g., Message Passing Interface (MPI) message sizes, kernel execution times, data movement rates) and dynamic resource topology (e.g., transient link failures, power fluctuations, active core counts). The SNN outputs predictive response times, resource saturation points, and potential thermal hotspots with probabilistic confidence levels. Based on these real-time predictions and pre-defined Quality-of-Service (QoS) objectives (e.g., maximum 5µs latency for critical path operations, 99.9999% throughput for specific data streams), an adaptive resource orchestrator, employing a hierarchical reinforcement learning agent with dynamic policy adaptation, allocates compute, memory, and high-speed interconnect fabric resources (e.g., InfiniBand, Slingshot) across the cluster. This orchestrator adjusts job scheduling policies (e.g., gang scheduling, topological scheduling), re-routes network traffic, and dynamically reconfigures memory placement in near real-time to mitigate predicted performance degradation and maintain hyper-scale efficiency.

graph TD
    A[Exascale Workload<br>(MPI, Kernels, Data Streams)] --> B(Distributed Telemetry Network<br>@100kHz+)
    B --> C{High-Frequency, High-Volume Performance Data}
    C --> D[Massively Parallel SNN Model<br>(GPU Clusters + NVLink)]
    D -- Predicted Response/Saturation/Hotspots --> E(Hierarchical RL Orchestrator<br>+ Dynamic Policy Engine)
    F[QoS Objectives<br>(Latency, Throughput, Efficiency)] --> E
    E --> G[Resource Allocation Actions<br>(Compute, Memory, Interconnect Routing)]
    G --> H[Exascale Supercomputing System]
    H -- Actual Micro-Latencies/Utilization --> B

3. Cross-Domain Application: Manufacturing/Industrial Automation

Enabling Description:
A method for adaptive resource allocation in a flexible manufacturing system (FMS) comprising a plurality of robotic workstations, CNC machines, and autonomous material handling units (e.g., AGVs, AMRs). Performance data, including robot cycle times, machine tool wear, buffer queue lengths, sensor fusion data (e.g., vision system defect rates, force sensor anomaly detection), and production throughput for different product families (job classes), is gathered via industrial IoT sensors communicating over OPC-UA and EtherCAT protocols. A deep neural network, specifically a Convolutional Neural Network (CNN) for sensor data processing combined with a Recurrent Neural Network (RNN) for temporal process dynamics, is constructed. This network has inputs representing machine states, current work-in-progress inventory levels, maintenance schedules, and production targets. The network is trained with historical operational data and simulated anomaly scenarios to model the FMS performance, predicting bottlenecks, quality deviations, and potential equipment failures. Based on these predictions and user-defined production objectives (e.g., maximize throughput for high-margin product A, minimize defect rate for critical component B, optimize energy consumption), a Manufacturing Execution System (MES) controller, integrating the neural network's real-time output, dynamically re-allocates resources. This includes adjusting robotic task assignments, re-prioritizing material flow via AGVs/AMRs, optimizing machine parameters (e.g., spindle speeds, feed rates, laser power), and dynamically adjusting buffer capacities to maintain optimal production flow and quality targets.

graph TD
    A[Production Orders/Workload] --> B(FMS IIoT Sensors<br>OPC-UA / EtherCAT)
    B --> C{Manufacturing Performance Data<br>(Sensor Fusion, Cycle Times, WIP)}
    C --> D[Hybrid CNN-RNN<br>(Production & Anomaly Model)]
    D -- Predicted Performance/Bottlenecks --> E(MES Controller<br>+ Optimization Engine)
    F[Production Objectives<br>(Throughput, Quality, Energy)] --> E
    E --> G[Resource Allocation Actions<br>(Robots, AGVs/AMRs, CNC, Buffers)]
    G --> H[Flexible Manufacturing System]
    H -- Actual Production Metrics/Feedback --> B

4. Cross-Domain Application: Smart Grid/Energy Management

Enabling Description:
A method for adaptive allocation of energy resources within a distributed smart grid infrastructure, balancing electricity demand and supply across various generation sources (e.g., solar farms, wind turbines, conventional power plants), energy storage units (e.g., grid-scale batteries), and demand-side management loads. Performance data, including real-time power generation output (MW), consumption patterns from different consumer classes (e.g., industrial, residential, commercial), dynamic grid stability metrics (e.g., voltage, frequency, phase angles, line losses), and energy storage levels (State of Charge), is gathered from Supervisory Control and Data Acquisition (SCADA) systems, smart meters, and grid sensors. A deep recurrent neural network (RNN), specifically a Long Short-Term Memory (LSTM) network, is constructed with inputs correlating real-time weather forecasts, historical demand profiles, fluctuating generation capacities, and dynamic grid topology. The LSTM network is trained to predict future demand-supply imbalances, grid stress points, and potential cascading failures across multiple look-ahead horizons. Based on these predictions and user-specified grid objectives (e.g., maximize renewable energy penetration, minimize curtailment, ensure N-1 security, optimize cost of energy distribution), a Grid Management System (GMS) controller, informed by the RNN's output, dynamically allocates energy resources. This involves real-time dispatching of generation units, scheduling battery charging/discharging cycles, enacting granular demand-response programs across consumer classes, and optimizing power flow control devices (e.g., Flexible AC Transmission Systems - FACTS) to maintain grid stability, efficiency, and resilience.

graph TD
    A[Energy Demand/Weather Forecast] --> B(SCADA/Smart Meters/Grid Sensors)
    B --> C{Real-time Energy Performance Data<br>(Generation, Consumption, Stability, Storage)}
    C --> D[Long Short-Term Memory (LSTM) Network<br>(Grid Prediction Model)]
    D -- Predicted Imbalances/Stress/Failures --> E(Grid Management System Controller<br>+ Optimization Algorithms)
    F[Grid Objectives<br>(Renewable % , Reliability, Cost)] --> E
    E --> G[Resource Allocation Actions<br>(Generation Dispatch, Battery Scheduling, Demand Response, FACTS)]
    G --> H[Distributed Smart Grid Infrastructure]
    H -- Actual Grid Metrics/Feedback --> B

5. Cross-Domain Application: Logistics/Supply Chain Optimization

Enabling Description:
A method for adaptive resource allocation in a global, multi-modal logistics and supply chain network, optimizing the movement and storage of goods across diverse transportation modes (ee.g., road, rail, air, sea) and interconnected warehouse/distribution centers. Performance data, including real-time vehicle utilization, estimated delivery times, container loading efficiencies, warehouse inventory levels, order fulfillment rates, port congestion, and traffic conditions for different product categories (job classes), is gathered from GPS trackers, telematics systems, warehouse management systems (WMS), and enterprise resource planning (ERP) systems. A Graph Neural Network (GNN) is constructed with inputs representing the dynamic supply chain network topology (nodes as locations, edges as routes), real-time demand fluctuations, fleet availability and status, weather conditions, and logistical constraints. The GNN is trained with historical logistics data, real-time sensor streams, and simulated disruption events (e.g., port strikes, vehicle breakdowns) to model complex supply chain dynamics, predicting delays, inventory shortages, optimal routing, and potential disruptions. Based on these predictions and user-specified logistics objectives (e.g., minimize transportation costs, maximize on-time delivery, optimize inventory turnover, reduce carbon footprint), a Supply Chain Orchestration (SCO) platform, utilizing the GNN's output, dynamically allocates resources. This includes optimizing fleet routing and scheduling, dynamically re-assigning warehouse picking and packing tasks, adjusting inventory placement across the network (e.g., pre-positioning goods), and re-negotiating carrier contracts in response to real-time conditions.

graph TD
    A[Orders/Demand Fluctuations] --> B(GPS/Telematics, WMS, ERP Systems<br>Real-time Sensor Data)
    B --> C{Logistics Performance Data<br>(Utilization, Delivery, Inventory, Traffic, Weather)}
    C --> D[Graph Neural Network (GNN)<br>(Supply Chain Dynamics Model)]
    D -- Predicted Delays/Shortages/Disruptions --> E(Supply Chain Orchestration Platform<br>+ Multi-Objective Optimizer)
    F[Logistics Objectives<br>(Cost, Delivery Speed, Inventory, CO2)] --> E
    E --> G[Resource Allocation Actions<br>(Fleet Routing, Warehouse Tasks, Inventory Placement, Carrier Mgmt.)]
    G --> H[Global Supply Chain Network]
    H -- Actual Logistics Metrics/Feedback --> B

6. Integration with Emerging Tech: Hybrid Cloud with AI/IoT/Blockchain

Enabling Description:
A method for self-optimizing, transparent, and auditable resource allocation in a hybrid cloud computing environment, integrating real-time IoT sensor data, advanced AI-driven meta-optimization, and blockchain-based provenance. Edge-deployed IoT sensors (e.g., power meters, environmental sensors, custom hardware probes) continuously monitor individual virtual machine (VM), container, and bare-metal server performance metrics (e.g., CPU cycles, memory pressure, I/O wait times, network latency, energy consumption, thermal profiles) and underlying infrastructure health. This fine-grained, high-frequency performance data is streamed to a central AI orchestration layer. A hierarchical ensemble of neural networks (e.g., a combination of deep feedforward networks for static configurations and recurrent networks for temporal patterns), orchestrated by a Reinforcement Learning (RL) agent, constructs a dynamic, predictive model of the hybrid cloud's behavior, predicting resource contention, Service Level Agreement (SLA) violations, and optimal power states. The RL agent, utilizing these predictions, dynamically adjusts resource allocations (e.g., VM migration across cloud providers, container auto-scaling policies, network bandwidth shaping, dynamic voltage and frequency scaling (DVFS)) across on-premise, private, and public cloud infrastructure. Each significant resource allocation decision, along with its justification derived from the neural network's probabilistic outputs and the RL agent's policy, is immutably recorded as a transaction on a permissioned distributed ledger (blockchain, e.g., Hyperledger Fabric). This blockchain ensures immutable auditability, transparency, and cryptographically provable compliance with user-specified performance objectives, contractual SLAs, and regulatory requirements. Furthermore, a Bayesian optimization algorithm continuously adapts and fine-tunes the neural network's hyperparameters and the RL agent's reward functions, achieving autonomous, meta-level optimization of the entire resource management system based on observed, long-term performance improvements and cost efficiencies.

graph TD
    A[Hybrid Cloud Workload & Traffic] --> B(IoT Edge Sensors + Software Performance Monitors)
    B --> C{Real-time Performance & Environmental Data}
    C --> D[Hierarchical NN Ensemble<br>(Cloud Behavior & SLA Prediction)]
    D -- Predicted Contention/Violations --> E(RL Orchestrator + Bayesian Optimizer)
    F[User SLOs/SLAs & Regulatory Compliance] --> E
    E --> G[Blockchain Ledger<br>(Immutable Allocation Records & Rationale)]
    E --> H[Dynamic Resource Allocator<br>(VMs, Containers, Network, DVFS)]
    H --> I[Hybrid Cloud Infrastructure<br>(On-prem, Private, Public)]
    I -- Actual Performance/Consumption --> B
    G -- Audit Trail/Verification --> J[Auditors/Users/Regulatory Bodies]

7. The "Inverse" or Failure Mode: Graceful Degradation & Safe-Mode Resource Allocation

Enabling Description:
A method for ensuring graceful degradation and safe-mode resource allocation in a mission-critical, real-time embedded system (e.g., an autonomous vehicle control unit, a nuclear power plant safety system, or a medical life-support device) operating under anticipated or actual component failures, environmental disturbances, or severe resource constraints. The system continuously gathers highly redundant performance data, including cross-checked sensor readings, internal diagnostic codes, CPU load, memory integrity checks (ECC), bus arbitration success rates, and task completion rates for safety-critical (high integrity) and non-safety-critical (low integrity) functions. A specialized, fault-tolerant neural network architecture (e.g., a modular neural network with active redundancy in critical layers, an ensemble of diverse neural networks with a voting mechanism, or a Bayesian Neural Network for uncertainty quantification) is constructed and trained. This training incorporates both normal operational data and an extensive library of simulated failure modes and degraded states. This neural network model continuously predicts potential system failures, resource exhaustion states, and deviations from a pre-defined "safe operating envelope" with probabilistic confidence. Upon detecting a predicted or actual failure/constraint (e.g., sensor malfunction, CPU core degradation, power supply instability, cyber-attack signature), the system automatically initiates a "limited-functionality" or "safe-mode" protocol. The neural network's output then directly guides a safety-critical resource manager, implemented as a certified real-time operating system (RTOS) kernel module, to:

1. Strictly prioritize safety-critical job classes: Allocate all available, verified minimal resources (e.g., 10% CPU, 20% memory, dedicated I/O channels) exclusively to core safety functions, aggressively shedding or suspending all non-essential services.
2. Activate redundant components and failover mechanisms: Trigger automatic switching to backup sensors, redundant processors, or failover network paths.
3. Initiate controlled shutdown or recovery procedures: Execute pre-defined, certified sequences for maintaining system integrity, safely transitioning to a minimal operational state, or preparing for external intervention.
   The allocation strategy ensures the system remains within pre-certified "safe-operating envelopes" by dynamically scaling back performance, shedding non-essential workload, and reconfiguring hardware based on the neural network's continuous, real-time assessment of system health, failure propagation, and available fault-isolated resources. The primary objective shifts from optimal performance to maximum safety and survivability.

graph TD
    A[Mission-Critical Workload<br>(Safety-Critical & Non-Critical)] --> B(Redundant System Sensors<br>+ Diagnostics & ECC)
    B --> C{Verified System Health & Performance Data<br>(incl. Anomaly Signatures)}
    C --> D[Fault-Tolerant Neural Network<br>(Failure Prediction & Uncertainty Quantification)]
    D -- Predicted Failure/Constraint & Confidence --> E{Operational Mode Decision<br>(RTOS Kernel Module)}
    E -- Normal Operation --> F[Normal Resource Allocation]
    E -- Safe Mode/Degradation --> G[Safety-Critical Resource Manager<br>(Prioritization, Redundancy Activation)]
    F --> H[Full System Functionality]
    G --> I[Limited Functionality/Safe Mode<br>(Core Safety Functions Only)]
    H -- Actual Performance Feedback --> B
    I -- Degraded Performance/Diagnostics --> B
    J[Safety Objectives<br>+ Failure Modes & Safe Operating Envelopes] --> D

---

Combination Prior Art Scenarios with Open-Source Standards

These scenarios demonstrate how the principles of US Patent 5704012 could be readily combined with existing, widely adopted open-source standards to create obvious improvements.

1. US5704012 + Linux Kernel & cgroups (Control Groups)

Description:
A PHOSITA would find it obvious to integrate the neural network-based adaptive resource allocation method of US5704012 with the cgroups (control groups) feature of the Linux kernel. The neural network controller, trained on performance data gathered from a Linux system running diverse workloads, would dynamically determine optimal resource parameters for different cgroups (e1.g., CPU shares, memory limits, I/O bandwidth, block I/O weights). The resource manager component of the operating system (as described in US5704012) would then translate these neural network outputs into specific cgroup configurations. These configurations would be applied in real-time by writing to the cgroup virtual filesystems (e.g., /sys/fs/cgroup/cpu/user_jobs/cpu.shares) to adjust the resource allocations for various job classes. This combination provides an intelligent, adaptive layer for managing system resources on top of a mature and widely adopted open-source Linux resource management framework, enabling the system to dynamically respond to changing workloads and configurations in an optimized manner.

2. US5704012 + Kubernetes Resource Management & Prometheus Monitoring

Description:
It would be obvious to integrate the adaptive neural network controller of US5704012 with Kubernetes for container orchestration and Prometheus for monitoring. Prometheus, acting as the "computer system performance monitor" (step 1 of Claim 1), would continuously collect granular metrics (e.g., CPU utilization, memory usage, network I/O, latency, error rates) from Kubernetes pods, nodes, and services. This performance data, including workload and configuration details (e.g., number of replicas, deployed services), would be used to train the neural network "system model" (step 3). The trained neural network "controller" (step 5) would then process user-defined performance objectives (e.g., target latency for a specific microservice, desired throughput for a batch processing job, maximum cost for a given workload) and output optimal resource requests, limits, and autoscaling parameters (e.g., minReplicas, maxReplicas for Horizontal Pod Autoscaler - HPA, or resource values for Vertical Pod Autoscaler - VPA). The Kubernetes API would then serve as the "resource manager" to effect these allocations by dynamically updating pod specifications, HPA/VPA configurations, or even influencing node scheduling decisions (e.g., taint/toleration, affinity rules) within the cluster. This combination enhances standard container orchestration with intelligent, adaptive resource management.

3. US5704012 + Apache Mesos/YARN & Hadoop/Spark Workloads

Description:
A PHOSITA would find it obvious to apply the neural network-based adaptive resource allocation method of US5704012 to cluster management frameworks like Apache Mesos or YARN (Yet Another Resource Negotiator) which manage distributed resources for big data processing frameworks such as Hadoop and Spark. Performance data, including job completion times, task execution durations, resource utilization per executor (CPU, memory, disk I/O), and network shuffle I/O for various Hadoop MapReduce, Spark Streaming, and Spark SQL job classes, would be gathered from the Mesos/YARN monitoring components. This data would be used to train a neural network to model the performance characteristics of the distributed cluster under different workloads and resource configurations. The neural network controller would then dynamically advise the Mesos/YARN scheduler or resource manager on optimal resource allocations for different job queues or frameworks. These recommendations would aim to meet user-defined performance objectives (e.g., guarantee 95th percentile latency for interactive Spark queries, complete daily Hadoop reports by a strict deadline, ensure fairness among multi-tenant workloads). The Mesos/YARN resource management APIs would be used to implement the neural network's recommendations, dynamically adjusting resource reservations, container sizes, and task priorities across the cluster.