Tongdao Think Tank 2024-01-08 16:54 Published in Beijing.
A new approach to natural language models is expanding AI applications in edge computing. Innovations in the use of natural language models have brought artificial intelligence to sensors that can be deployed in the field, including drones.
Los Alamos National Laboratory in the United States is exploring artificial intelligence technology to locate and characterize abandoned oil and gas wells that emit climate-warming methane, according to Science & Technology on January 7.
Advanced artificial intelligence (AI) technology can reconstruct extensive data sets, such as total ocean temperatures, using a minimal number of sensors placed in the field. This approach leverages energy-efficient "edge" computing to offer a wide range of potential uses in various fields such as industry, scientific research, and healthcare.
"We've developed a neural network that allows us to represent large systems in a very compact way," said researchers at Alamos National Laboratory. This compactness means it requires fewer computing resources than state-of-the-art convolutional neural network architectures, making it ideal for field deployments of drones, sensor arrays, and other edge computing applications, bringing computing closer to the end point of use. ”
Novel AI approaches improve computational efficiency. J**ier Santos' research work builds on an artificial intelligence model developed by Google called Perceiver IO, applying the technology of natural language models such as ChatGPT to the problem of reconstructing information about vast areas, such as the ocean, from relatively few measurements.
The team at J**ier Santos realized that the model had a wide range of applications due to its high efficiency. Santos and his Los Alamos colleagues validated the model for the first time, demonstrating its effectiveness for real-world sparse datasets (i.e., information obtained from sensors that cover only a small portion of the area of interest) and complex datasets.
To demonstrate the real-world utility of the senseiver, the team applied the model to the National Oceanic and Atmospheric Administration's sea surface temperature dataset. The model is capable of integrating a large number of measurements obtained from satellites and onboard sensors over decades. Based on these sparse point measurements, the model can ** the temperature of the entire ocean, which provides useful information for global climate models.
"Alamos has a wide range of remote sensing capabilities, but it's not easy to use AI because the model is too big for field devices, which led us to turn to edge computing," said Hari Viswanathan of Alamos National Laboratory. ”
The approach provides improved capabilities for large-scale real-world applications, such as autonomous vehicles, remote modeling of oil and gas assets, medical monitoring of patients, cloud gaming, content delivery, and contaminant tracking.