Machine learning for thermal energy storage charging and dispatch
Optimal dispatch strategies for combined solar and battery systems
AI optimization of compressed air energy storage systems
Battery operators need to schedule storage charging and discharging under uncertain market conditions; prediction errors can reduce arbitrage value and lead to suboptimal dispatch.
Nuclear operators need to prepare for rare, high-risk emergencies where manual scenario planning is too slow and limited. Battery operators need dispatch decisions that maximize storage value under volatile power prices and system constraints; forecast-only approaches can miss the best control actions. Commercial deployment is hindered by poor model transferability, privacy concerns, and lack of trust in black-box models, especially in distributed and safety-critical storage environments.
Detects and diagnoses distribution network faults while improving circuit balancing and DER-aware grid management through visibility into changing load patterns and intermittent distributed energy resources.
LLM-based evaluation platform for credit-scoring and financial-analysis responses, automating open-ended answer grading at scale while aligning closely with human judgment.
Machine learning systems for optimizing battery storage dispatch, state of charge management, and grid-scale energy storage operations.
Machine learning for hydrogen storage management and optimization
It maximizes profits and reduces risks in hydrogen production and management. It optimizes hydrogen production and storage to reduce costs and improve efficiency. Hydrogen plants using scheduled or reactive maintenance face unnecessary downtime, higher maintenance costs, and lower reliability because failures are often addressed too late.
AI-powered object detection models analyze multi-source satellite, aerial, and SAR imagery to identify, classify, and track military and maritime assets in real time. By automating wide-area monitoring, change detection, and dark or disguised vessel discovery, it delivers faster, more accurate geospatial intelligence. Defense organizations gain earlier threat warning, improved mission planning, and more efficient use of ISR and analyst resources.
This AI solution uses AI to optimize inventory storage, warehouse operations, and end-to-end supply chain flows in manufacturing. It combines predictive logistics, real-time visibility, and autonomous warehouse robotics to minimize stockouts, excess inventory, and handling time. Manufacturers gain higher throughput, lower working capital, and more resilient, responsive supply networks.
Uses AI to optimize charge/discharge decisions under price uncertainty to maximize market and tariff value while managing degradation.
AI-driven optimization of data center cooling, power distribution, and energy efficiency.
Manual inspection in radioactive areas is slow, risky, and prone to human error. Helps facilities balance EV charging demand, storage usage, and local energy objectives to improve autonomy and reduce grid dependence. Reduces instability caused by fluctuating renewable generation and helps maintain reliable electricity delivery.
AI-driven optimization of flow battery systems
Object detection is a computer vision technique that simultaneously identifies what objects are present in an image or video frame and where they are located. It outputs bounding boxes (or sometimes masks), class labels, and confidence scores for each detected object. Modern approaches use deep learning architectures such as convolutional neural networks and vision transformers to learn visual features and regress object positions. Object detection is a core building block for perception in robotics, autonomous systems, and many real-time analytics applications.
Instance segmentation is a computer vision approach that detects every object in an image and assigns a separate pixel-accurate mask to each individual instance, even when multiple objects share the same class. It combines object detection (localizing and classifying objects) with semantic segmentation (labeling pixels) to produce fine-grained, per-object shapes. Modern systems typically use deep neural networks to jointly predict bounding boxes, class labels, and masks in one or multiple stages. This enables precise reasoning about object boundaries, overlaps, and counts in complex scenes.
Cloud/Storage
Other
Other
Other
Other
Other
Other
Other
Other
Other
Other
Other
Other
Other
Other
Other
Other
Other
Other
Other
Other automotive SoC storage safety frameworks appears in 1 scoped applications and is modeled as a canonical company.