Modeling & Simulation for Smart Grids & Networks

FLX AI can create models using machine learning that simulate systems such as energy grids, or transportation networks. Using simulated “what if” scenarios we can test a system’s capability to handle potential interactions, such as an increase in load upon an electrical grid, without disruption to the actual network in operation.

Power Grid Simulation

FLX AI can simulate interactions that occur when intermittent generation methods like wind and solar are added to the grid or when the system incurs a large power draw such as the addition of a large campus. We apply active learning methodologies to continually update the model with new information.

Demand Forecasting for Shared Services

For services such as bike or car share, we can create models that anticipate changes in demand so as to enable optimal distribution of inventory.

Resource Allocation

We can construct models that illustrate how to apply limited human or material resources so as to achieve optimal outcomes.

Business Intelligence

We pull from reservoirs of retail data — point-of-purchase, psychographic and demographic — to provide critical insights that support revenue growth, mitigate risk in capital investments, and diminish customer churn.

Marketing Mix Models

We can provide insights that enable you to clearly see which communications channels are most effective in driving revenue. Our data-driven approach will enable you to maximize return on investment for your marketing spend.

Customer Segmentation

We develop data-driven insights about your customer base to determine attributes of high-value customers. We then use those insights with individual-level databases to drive targeted marketing campaigns.

Retail Sales Optimization

We can perform market basket analyses that serve as the basis for upselling, cross selling and shelving strategies. Through data analysis, we can determine the probability that a customer who bought one item will purchase another or support the development of pricing strategies that maximize revenue.

Customer Churn Analysis

We apply data science to determine what matters most to your high-value customers. We then identify the actionable indicators that enable you to mitigate churn.

Demand Modeling Algorithms

If you’re looking to expand your business or franchise, we can identify and characterize your customer base, then determine the best areas and customer segments toward which to target marketing resources. Our data-driven approach can help you to decrease the risk of capital-intensive campaigns.

Image & Video Analytics

FLX AI possesses considerable breadth and depth of image and video analytics expertise. We can apply artificial intelligence and machine learning to provide automated solutions that perceive patterns, classify objects, and interpret events with speed and precision, saving time and money and eliminating human error.

Medical Imaging Analytics

Medical imaging analytics requires sophisticated techniques to understand and classify X-Rays, MRIs, and micro-video imagery. In a hospital or clinical research setting, imaging analytics accelerates throughput with automated analysis that reduces exponentially the images (MRIs or X-Rays) that a radiologist must review and improves repeatability or precision by eliminating subjectivity. AI can be applied to determine image quality automatically, saving patients and providers the time and expanse of having to repeat imaging. In telemedicine AI solutions could enhance efficiency by providing immediate feedback on image quality.

Surveillance & Security

With AI and machine learning, FLX AI can interpret satellite imagery, segmenting stationary and moving objects to analyze a scene and report on events. Traffic trends and buying patterns can be analyzed from images of streets and parking lots. Anomalies that may represent security threats can be identified.

Predictive Maintenance & Quality Control

FLX AI can apply analytics for continuous defect detection or predictive condition-based maintenance where computer vision is applied to inspect parts. Cameras can be used in place of shift workers to perform quality control in manufacturing lines.

Sensors, Wearable Devices, & the Internet of Things (IoT)

Connections among devices, sensors, and other items embedded with electronics in what’s known as the Internet of Things has enabled new applications in areas such as health care and life sciences. Data from multiple sources must be gathered, processed and understood in real time to take advantage of the information these networks can provide. FLX AI is experienced in handling issues that can arise with these data systems. For instance, we apply imputation techniques when data is lost or of poor quality and ensure records remain balanced so results are robust and unbiased. To keep solutions transparent, we implement Explainable AI to reveal the features that influence the classification of things.

Wearable Medical Devices

Medical devices must continuously monitor a multitude of biological and environmental signals that require significant integration and analysis of data. FLX AI researchers, applying various machine learning approaches, fuse the data from each sensor in order to understand the presence of medically relevant events and predict the onset of acute events in the future.

Passive Sensors

Patients post-surgery or with acute conditions often fail to comply with their physician’s requests for monitoring and tracking. FLX AI has worked on passive monitoring solutions that solve the problem through interpreting multiple signals from devises, essentially monitoring the patient’s condition invisibly and providing physicians with critical information.

Geospatial Imaging & Image Forensics

Applying our expertise in computer vision and machine learning, FLX AI can work with you to interpret and authenticate satellite imagery. We offer solutions that work for defense contractors, government agencies and a wide range of companies looking to identify patterns and trends and gain real-time intelligence from images.

Image Understanding

FLX AI uses imaging data from satellite and drone technology to understand in real time the implications of activities seen on the ground. We apply multimodal data analysis and fusion with data sources that include satellite and wide area motion imagery (WAMI), open source maps and web pages.

Image Forensics

When a government turns to the private sector for satellite and drone data it loses control over data provenance and must have means to identify signs of malicious tampering or manipulation in both publicly and privately available satellite data. Satellite meta-data about images and videos can also be hacked. We apply data fusion and computer vision techniques such as supervised deep learning to detect inconsistencies and anomalies in both image content and its descriptive metadata.

Platforms & Data Management

FLX AI works with organizations that have a variety of data management issues to resolve. Data overload can cause systems to run slowly. Or data coming from diverse sources may need to be converted to a consistent format before it can be analyzed. FLX AI creates automated platforms optimized for rapid analysis which can be stored on premises, deployed on the cloud, or a combination of both. Once a system is running, we provide DevOps to maintain it with software upgrades and monitoring to prevent downtime.

Data Integrations

We work with clients to bring together data of different types and from different sources into one set that can be analyzed to produce meaningful insights. Health care data from various providers, for instance, must be processed to enable analysis so public health officials can make data-driven policy decisions. Or data from different departments in an organization – finance, HR, marketing – may need to be integrated and analyzed before it can be expressed in a dashboard that provides insights and predictions for management.

Real Time Analysis

FLX AI can provide stream analytics, analyzing data as it is moving to provide real time dashboards that monitor health, traffic congestion, and any number of events that benefit from real-time analysis.