Artificial intelligence and machine learning technologies have been applied in a wide range of industries and domains, including healthcare, finance and governance, sports and entertainment, mining, construction, logistics, food, fashion, and many others. AI and ML technologies have been notably successful in automation of digital and physical tasks using automated workflows and robotics technology, respectively. They have also been instrumental in deriving insights from the huge amounts of text and numeric data being produced by various systems using natural language processing and text mining tools. Further, AI and ML have begun to play an enormous role in decision-making support through the development of tools using summarization, information retrieval, and data mining and techniques to predict (and prevent or drive) various outcomes and events using classification, regression, and clustering techniques.
One area in particular that is gaining traction is AI in anomaly detection. Anomaly detection has been successfully applied to optimize operations in a number of industries. The anomaly detection market is expected to reach $4.45 billion by 2023.
What is anomaly detection?
Anomaly detection aims to identify unusual patterns, anomalies, or data points that do not conform to the expected distribution. Applications of anomaly detection include fraud detection in financial transactions, fault detection in manufacturing, intrusion detection in a computer network, monitoring sensor readings in an aircraft, spotting potential risk or medical problems in health data, and predictive maintenance.
AI-driven anomaly detection for construction productivity:
Doxel is a Silicon Valley startup providing AI-enhanced software focused on improving construction productivity. Doxel uses rugged robots and drones equipped with cameras and LiDAR sensors to monitor and scan worksites. A Doxel robot scans construction sites every day to monitor how things are progressing, tracking what gets installed and whether it’s the right equipment, at the right time, in the right place. Once a construction site shuts down for the night, the small robot deployed by Doxel can get to work. The robot scans the site and uploads data to the cloud. There, deep learning algorithms flag anything that deviates from the building plans so that a manager can fix it the day after. The robot has no problem following prescheduled paths that can include stairs, and just one of these robots can scan about 30,000 square meters over the course of a week.
The visual data collected by their robot is also processed to measure the currently installed units and the rate of production by matching acquired data against the desired planning and design parameters for the client. The company states that their AI platform can also detect errors in construction by comparing visual data from daily scans of the job site to small-scale design models. Doxel has had successful collaborations in the past, such as their project for the Kaiser Permanente Viewridge Medical Office project. Their real-time progress-tracking system was able to prompt the Viewridge Medical Office project team to take action when predefined schedule deviations were detected, eventually yielding a 38% increase in labor productivity across all teams involved in the construction, thus enabling the project to be completed 11% under budget.
AI-driven anomaly detection for manufacturing:
General Electric launched its Brilliant Manufacturing Suite for customers, which the company had been field testing in its own factories. The system takes a holistic approach of tracking and processing everything in the manufacturing process to find possible issues before they emerge and detect inefficiencies. Their first “brilliant factory,” a $200-million investment, was built in 2015 in Pune, India.
AI-driven anomaly detection for energy optimization:
NextEra Energy software, developed by SpaceTime Insight (now part of Nokia), helps with performance optimization, real-time diagnostics and troubleshooting, and maintenance crew scheduling. NextEra’s ControlComm is built on a communications platform developed by AutoGrid, which helps business customers reduce energy bills during times of peak energy demand or high wholesale electricity prices by adjusting their energy consumption manually or with an automated solution.
Duke Energy developed the SmartGen program, which saved them more than $4.1 million by triggering an early warning when one of Duke Energy’s steam turbines had begun to malfunction. Duke Energy also uses SparkCogntition’s AI algorithms to safeguard its multibillion-dollar turbines by predicting potential disasters and shutdowns.
SLAC National Accelerator Laboratory uses AI to optimize solar and other distributed energy resources. Recent news reports describe new research efforts from the US Energy Department’s SLAC National Accelerator Laboratory at Stanford University that use AI applications to help utilities better integrate their solar resources and make more informed planning decisions for enhancing grid reliability, resiliency, and security.
This excerpt was taken from our Disruptors report titled “AI and Industry: How Machine Learning Is Impacting Industries.” The full report can be viewed here.
