AI Sustainability Control for Cloud Data Centers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional sustainability control in cloud-based data centers is typically static and separate from functional objective performance evaluation, making it difficult to dynamically manage and optimize environmental sustainability alongside processing reliability and efficiency.
Innovation Solution
A computer-implemented method using artificial intelligence to learn dynamic key performance indicators that relate functional objective performance to sustainability, identifying anomalies, and generating corrective actions to autonomously remediate risks, thereby enhancing sustainability control and system processing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional static key performance indicators are used for sustainability control, then the control process is simple and easy to implement, but the system cannot dynamically adapt to changing demands and goals
Solution Approach 1:
The patent applies dynamics by transitioning from static key performance indicators to dynamic KPIs that automatically adapt to changing demands and goals. The AI model continuously learns and updates KPIs based on real-time data, enabling the sustainability control process to respond dynamically to varying conditions without manual reconfiguration.
Solution Approach 2:
The system implements self-service through autonomous AI-based learning and adaptation. The AI model automatically learns from heterogeneous data sources, identifies patterns, and updates key performance indicators without human intervention. This enables the system to self-adjust to changing demands while managing complexity through automation.
2Reliability
If sustainability control is separated from functional objective performance evaluation, then the evaluation processes can be simplified and focused, but it becomes difficult to manage and optimize environmental sustainability alongside processing reliability
Solution Approach 1:
The patent merges sustainability control with functional objective performance evaluation by integrating both into a unified AI-based monitoring system. The system simultaneously evaluates processing reliability metrics and environmental sustainability indicators, allowing for coordinated optimization of both objectives through a single integrated framework.
Solution Approach 2:
The AI model serves multiple functions by handling both reliability evaluation and sustainability assessment within a single system. It processes heterogeneous data from various sources to generate unified insights that address both processing performance and environmental impact, eliminating the need for separate specialized systems.
3Adaptability or versatility
If heterogeneous data from multiple data sources is collected and normalized, then comprehensive sustainability control can be achieved, but the data processing complexity increases
Solution Approach 1:
The patent replaces manual data processing mechanisms with AI-based automated systems. Instead of requiring complex manual normalization and integration of heterogeneous data, the AI model automatically learns patterns, identifies relationships, and processes data from multiple sources, reducing processing complexity through intelligent automation.
Solution Approach 2:
The system performs self-service data processing by automatically learning from heterogeneous data sources and normalizing data formats without human intervention. The AI model autonomously handles data collection, classification, and integration, managing processing complexity through self-organized learning and adaptation.
4Productivity
If AI models are trained to learn dynamic key performance indicators, then sustainable optimization can be achieved, but the training time and computational resources required increase
Solution Approach 1:
The patent applies preliminary action by pre-training AI models on historical data and establishing baseline KPI relationships before actual sustainability optimization is needed. This preliminary training enables the model to quickly adapt to new demands without requiring extensive retraining, reducing overall training time while maintaining optimization efficiency.
Solution Approach 2:
The system uses dynamic learning where the AI model continuously updates its knowledge based on new data rather than requiring complete retraining. This dynamic adaptation allows the model to maintain high optimization efficiency while minimizing training time by leveraging previously learned patterns and relationships.
Data Source
AI summary
Autonomous sustainability control is provided related to performing a functional objective. Normalized data is obtained from heterogeneous data obtained from a plurality of data sources, where the heterogeneous data is related, at least in part, to the functional objective. The normalized data is used to train an artificial intelligence model to learn dynamic key performance indicators relating, at least in part, performance of the functional objective to sustainability. The artificial intelligence model is used to learn a set of dynamic key performance indicators to relate current performance of the functional objective to sustainability. The learned set of dynamic key performance indicators is used to identify an anomaly, and one or more corrective actions are generated to remediate a risk associated with the anomaly. The one or more actions facilitate the autonomous sustainability control related to performing the functional objective.