Researchers are now closer to solving a long-standing challenge in earthquake prediction science. Scientists at Penn State University enhanced an advanced AI system for monitoring laboratory-created earthquakes known as “labquakes.” Scientific teams now plan to perform lab tests to teach us about earthquake mechanics and help us forecast future earthquakes.
Scientists create earthquake simulations by sliding blocks of rock; known as friction experiments, under carefully monitored stress at controlled levels. This allows the tracking of important physical and mechanical factors that contribute to fault failure outcomes through this testing design. Laboratory-created earthquakes let scientists study precise details of how rock stress affects fault reactions because their monitoring instruments are close to the fault zone.
“We are a long way off from predicting natural earthquakes, but understanding the physics of labquakes and how they evolve allows us to better understand the mechanics of real earthquakes,”
stated Parisa Shokouhi, the Penn State professor of engineering science and of acoustics.
Research works best when scientists study how minor rock fragments affect fault movement inside controlled laboratory conditions. The research gives scientists important knowledge about the specific events that force stable faults to transition into failure mode.
Parisa Shokouhi and her team developed a machine learning model for labquake prediction that can also automatically retrieve specific parameters, which are known as rate and state friction parameters, from the ultrasonic monitoring of ‘labquake’ experiments. The rate and state friction parameters define the mechanics of the labquakes; they determine the strength of the fault, signalling how close it is to failure.
“We can study the precise conditions under which a gently creeping fault suddenly becomes unstable and triggers an earthquake, such as the amount of stress, the roughness of the fault or the role of small loose rock particles at the interface, to name a few possibilities.”
To estimate these parameters, the team developed a physics-informed neural network (PINN). The PINN learns better results by applying physical rules restrictions while preparing itself to perform under limited available data. Having this ability gives PINN an edge over traditional data-driven models that need big datasets and have problems extending to new scenarios.
Through transfer learning PINN finds out how understanding of one environment impacts another. Laboratory and field applications in seismology need to connect better through physical inverse neural networks.
“We show that PINN models provide accurate predictions with a smaller amount of training data and that transfer learning—when trained models are applied to a new, related task—is greatly enhanced in these models,”
stated Jacques Rivière, co-author and assistant professor of engineering science and mechanics at Penn State.
The development and training of the physics-informed neural networks, Prabhav Borate, a graduate student in engineering science, used labquake data collected in the Rock Mechanics Laboratory of co-author Chris Marone, professor of geosciences in the College of Earth and Mineral Sciences.
“We created the PINN models by training them to follow the rate and state friction laws,”
Borate Stated. He went onto say:
“This was achieved by designing the model to penalise itself whenever the predictions didn’t match the law. This approach proved effective in accurate prediction of labquakes using smaller datasets while providing invaluable information about the earthquake mechanics through the extracted friction parameters.”
While predicting natural earthquakes remains a distant goal, this research lays the groundwork for more comprehensive models that incorporate both physical laws and real-world data. The ability to monitor and predict fault behaviour with high precision could lead to significant advancements in seismic hazard assessment and risk mitigation strategies.

Hassan graduated with a Master’s degree in Chemical Engineering from the University of Chester (UK). He currently works as a design engineering consultant for one of the largest engineering firms in the world along with being an associate member of the Institute of Chemical Engineers (IChemE).