John Taylor, Professor of Economics at Stanford University and developer of the "Taylor Rule" for setting interest rates | Stanford University
John Taylor, Professor of Economics at Stanford University and developer of the "Taylor Rule" for setting interest rates | Stanford University
Stanford researchers have developed an artificial intelligence-based tool, SandAI, that can reveal the history of quartz sand grains dating back hundreds of millions of years. SandAI allows researchers to determine with high accuracy whether wind, rivers, waves, or glacial movements shaped and deposited sand grains.
The tool provides a unique window into the past for geological and archaeological studies, particularly for periods and environments where few other clues, such as fossils, are preserved. SandAI’s approach, known as microtextural analysis, also aids modern forensic investigations into illegal sand mining.
“Working on sedimentary deposits that haven’t been disturbed or deformed feels about as close as you can get to being in a time machine – you’re seeing exactly what was on the surface of Earth, even hundreds of millions of years ago. SandAI adds another layer of detail to the information we can pull from them,” said Michael Hasson, a PhD candidate with Mathieu Lapôtre at the Stanford Doerr School of Sustainability. Hasson is the lead author of a new study demonstrating the tool, published this week in Proceedings of the National Academy of Sciences.
Historically, microtextural analysis has been performed manually using magnifying glasses and microscopes to infer sand grains’ histories. Modern science has validated this approach by showing that transport mechanisms leave telltale signatures—grains traveling farther often appear more rounded due to abrasion patterns from waves and wind.
However, traditional microtextural analysis is subjective and time-consuming. The new tool leverages machine learning to scrutinize microscopic images of sand grains quantitatively and objectively. It analyzes individual sand grains instead of grouping multiple grains into single categories for a more comprehensive evaluation.
“Instead of a human going through and deciding what one texture is versus another for sand grains, we are using machine learning to make microtextural analysis more objective and rigorous,” said Lapôtre. “Our tool is opening doors for microtextural analysis applications that were not available before.”
Sand is the most used resource worldwide after water and is critical in construction materials like concrete and mortar. Ensuring ethical sourcing is challenging; hence researchers hope SandAI can enhance traceability and aid forensic investigations into illegal sand mining.
To build SandAI, researchers employed a neural network mimicking human brain functions to learn from mistakes. With global collaborators' help, Hasson assembled hundreds of scanning electron microscope images representing common terrestrial environments: fluvial (rivers), eolian (windblown sediments), glacial, and beach.
“We wanted this method to work across geological time but also across all geography on Earth,” said Hasson. “For example, the windblown dunes class included wet and dry examples.”
SandAI trained itself on these images to predict sand grains’ histories based on features humans might not discern. Once achieving 90% prediction accuracy with known samples ranging up to 200 million years old from well-characterized environments like the Jurassic era, it analyzed older samples dating back over 600 million years from Norway's Cryogenian period.
“With this Cryogenian sample, we were seeing how far we can push SandAI,” Hasson said.
SandAI suggested ancient sand grains had been part of windblown dunes near glaciers during Snowball Earth—a conclusion aligning with some manual studies but providing additional details missed by traditional techniques.
To further evaluate findings, researchers compared results with modern Antarctic windblown sands using SandAI—confirming Antarctica as an analog environment for Cryogenian deposits.
“These findings suggest Antarctica really is a good modern analog to the environment represented by the Bråvika Member,” Hasson said. “They are strong evidence that our signal from Cryogenian deposits isn’t just a fluke.”
The researchers have made SandAI available online for public use and plan ongoing development based on user feedback.
“The fact that we can now offer detailed conclusions about geological deposits that weren’t knowable before I find kind of mind-blowing,” said Hasson. “We’re looking forward to seeing what else SandAI can do.”
Media contacts:
Mathieu Lapôtre: 626-232-5494
Michael Hasson: 415-250-1799
Danielle Torrent Tucker: 650-497-9541
©Copyright Stanford University
###
Alerts Sign-up