AI Strips out Noise

Note AI methods in use for noice removal. 

Deep Learning, Supervised Learning. Other applications in noise reduction?

AI Strips Out City Noise to Improve Earthquake Monitoring Systems

New Scientist, Chris Stokel-Walker, April 13, 2022

Stanford University's Gregory Baroza and colleagues used a deep learning algorithm to eliminate city noise from earthquake monitors, in an attempt to fine-tune the ability to locate where tremors originate. The researchers trained the artificial intelligence on 80,000 samples of urban noise and 33,751 samples of earthquake signals to distinguish between the two. Running audio through the neural network enhanced the signal-to-noise ratio by an average of 15 decibels, triple the average of previous denoising methods. Rice University's Maarten de Hoop said one shortcoming of the approach was the network's training via supervised learning using human-labeled data sampled from one area; he said this makes the technique less likely to be effective when presented with noise from somewhere else. ... '

ShakeAlert Earthquake Warning System

 Were loosely involved in the testing of this warning system. Up to 10 second pre alert noted.

Home/News/Seconds Before Earthquake Rattled California, Phones... 

ACM TECHNEWS

Seconds Before Earthquake Rattled California, Phones Got Vital Warning  By The Guardian (U.K.)

An early-alert system managed by the U.S. Geological Survey (USGS) on Monday warned Californians of a 6.2-magnitude earthquake by phone, seconds before it struck.

The ShakeAlert system issues warnings through various agencies and applications, including Google's Android operating system.   USGS sensors feed information bundled into a data package that is displayed on phones within seconds; some alert apps are available for download, but even some who had no such app on their phones received alerts.

ShakeAlert sent quake warnings to about 500,000 phones before the tremors began.  "We got some reports from folks that they got up to 10 seconds' warning before they felt shaking. That's pretty darn good," said the USGS' Robert de Groot.

From The Guardian (U.K.)

View Full Article  

See also:

Using Sparse Data to Predict Lab Quakes  By Los Alamos National Laboratory,  December 30, 2021

Better Detection of Earthquakes with Machine Learning

New work in the space using Machine Learning, we had proposed related methods when studying neural models.  Here a convolutional neural network.

Researchers Create Earthquake System Model with Better Detection Capabilities  in CACM

University of Wyoming,  October 12, 2021

The University of Wyoming's Pejman Tahmasebi and Tao Bai have invented a machine learning (ML) model that boosts the accuracy of earthquake detection significantly over current models. Tahmasebi said the model processes signal data recorded by seismometers, and can automatically distinguish seismic events from seismic noise. The model combines existing long short-term memory and fully convolutional network ML models; the former captures data signal changes over time, and the latter filters out hidden features of seismic events. Tahmasebi said the model boasts 89.1% classification accuracy, a 14.5% improvement over the state-of-the-art ConvNetQuake model.

Full article

NVIDIA Predicting Earth Quake Intensity

Predicting earthquake and their intensity was something we proposed some years ago, mentioned here.  glad to see the idea taken much further.    Looking at this application further. 

AI of Earthshaking Magnitude: DeepShake Predicts Quake Intensity   By Isha Salian

Tags: Deep Learning, featured, Geoscience, News

In a major earthquake, even a few seconds of advance warning can help people prepare — so Stanford University researchers have turned to deep learning to predict strong shaking and issue early alerts.

DeepShake, a spatiotemporal neural network trained on seismic recordings from around 30,000 earthquakes, analyzes seismic signals in real time. By observing the earliest detected waves from an earthquake, the neural network can predict ground shaking intensity and send alerts throughout the area. 

Geophysics and computer science researchers at Stanford used a university cluster of NVIDIA GPUs to develop the model, using data from the 2019 Ridgecrest sequence of earthquakes in Southern Califonia. 

When tested with seismic data from Ridgecrest’s 7.1 magnitude earthquake, DeepShake provided simulated alerts to nearby seismic stations 7 to 13 seconds before the arrival of high intensity ground shaking.

Most early warning systems pull multiple information sources, first determining the location and magnitude of an earthquake before calculating ground motion for a specific area. 

“Each of these steps can introduce error that can degrade the ground shaking forecast,” said Stanford student Daniel Wu, who presented the project at the 2021 Annual Meeting of the Seismological Society of America. 

Instead, the DeepShake network relies solely on seismic waveforms for its rapid early warning and forecasting system. The unsupervised neural network learned which features of seismic waveform data best forecast the strength of future shaking. 

“We’ve noticed from building other neural networks for use in seismology that they can learn all sorts of interesting things, and so they might not need the epicenter and magnitude of the earthquake to make a good forecast,” said Wu. “DeepShake is trained on a preselected network of seismic stations, so that the local characteristics of those stations become part of the training data.”  ... ' 

Detecting Earthquakes and Waves

A favorite topic, detecting pattern.  Its a big part of our intelligence.   Here using new kinds of sensor data. 

Google Uses Underwater Fiber-Optic Cable to Detect Earthquakes   By New Scientist

A submarine fiber-optic cable owned by Google was used by researchers at the search engine giant and California Institute of Technology (Caltech) to detect earthquakes and ocean waves generated by storms.

The investigators measured changes in pressure and strain using traffic data from the 10,000-kilometer (6,213-mile)-long cable on the floor of the Pacific Ocean, recording about 30 ocean storm swell events and roughly 20 quakes exceeding magnitude 5 over nine months. Caltech's Zhongwen Zhan described this approach as more flexible and scalable than other attempts to deploy fiber-optic sensors, as new infrastructure is unnecessary.   Anthony Sladen at the University Côte d’Azur in France says the study constitutes “a major step in exploiting the benefits of existing cables.”

From New Scientist

Can AI Find Evolving Earthquakes?

More advances moving towards closer to rea-time prediction?

Technology Finds Long-Hidden Quakes, Possible Clues About How Earthquakes Evolve

Stanford News, Josie Garthwaite

Stanford University scientists developed new algorithms that extract evidence of long-hidden microquakes from massive seismic datasets. The Earthquake Transformer algorithm emulates how human analysts holistically analyze a set of seismic "wiggles," then focuses on a small section of interest. The Stanford team measured the algorithm’s performance using five weeks of data recorded in the region of Japan impacted two decades ago by the Tottori earthquake and its aftershocks. The algorithm detected 21,092 events—more than 2.5 times the number of quakes detected manually—within 20 minutes, using data from just 18 of 57 stations originally used to study the sequence. Said Stanford’s Gregory Beroza, “Earthquake monitoring using machine learning in near-real time is coming very soon.”

AI and Seismology

 An area we consulted on early and continue to follow.   Some comment here on why current, otherwise excellent pattern recognition techniques do not predict earthquakes.  Good update by CACM

AI Shakes up Seismology World

Commissioned by CACM Staff    By Samuel Greengard  

Artificial intelligence is helping researchers to better understand seismic events and to develop early warning systems that can save lives and protect property.

Few events on our planet are as complex as earthquakes. How, when and why they occur remains mostly a mystery, even with today's sophisticated instruments, sensors, and machines continuously monitoring and measuring seismic activity. "The vast number of variables and data points produce an extraordinarily complex picture," says Men-Andrin Meier, associate staff seismologist in the Seismological Laboratory of the California Institute of Technology (CalTech).

For decades, scientists have attempted to understand earthquakes using everything from satellite imagery to computer simulations, which have yielded modest and mixed results. Now scientists are turning to a new ally: artificial intelligence (AI), which is helping researchers better understand seismic events and develop early warning systems that can save lives and protect property.

"Machine learning and other forms of AI have emerged as valuable tools. They are advancing the science in a significant way," says Zachary Ross, an assistant professor of geophysics in the Division of Geological and Planetary Sciences at CalTech.

Finding Faults

The science surrounding earthquakes is extraordinarily complex. Unlike weather forecasting, which uses real-time data from satellites, sensors, and earth stations to track conditions as they occur, seismologists must rely on signals after an event. This data streams in from digital seismometers and broadband sensors on the ground. Measuring stress beneath the Earth's surface is next to impossible, because researchers don't have access to sensors buried deeply enough to measure underground forces.

Seismologists had largely given up on the idea of predicting earthquakes; at least, for the foreseeable future. However, the field is enjoying a renaissance thanks to machine learning and deep learning. Using connected sensors and algorithms, researchers are gaining insights into earthquake behavior, including how smaller swarms of temblors may or may not lead to a larger event. The researchers also are developing early alert systems that can protect property and lives. Says Meier, "We have gotten to the point where we don't have to choose between quantity and quality of the data."  .... " 

Phones Create Earthquake Monitor

Saw similar applications suggested for some time.  I am still a member of a CA based system that in part combines sensors and crowd sourcing monitors.  I get alerts weekly.  And will be alerted if people need to take quick action.  Now Google is seriously in the loop.

Google turns Android phones into an earthquake detection network
Your phones are essentially mini seismometers.
By Cherlynn Low, @cherlynnlow  in Engadget

In a natural disaster like an earthquake, even a few seconds’ heads up could save lives. You could use that time to get you and your loved ones somewhere safe and prevent fatalities and injuries. Google is rolling out a feature today that not only gives you an early warning about potential earthquakes happening, but also turns your phone into a mini seismometer and makes it part of a network of shockwave-detecting Android devices.

“The public infrastructure to detect and alert everyone about an earthquake is costly to deploy,” Google’s principal software engineer of Android Marc Stogaitis said in a blog post today. Not everyone is signed up to receive text alerts, for example, so if the early warning system is built into the OS, these messages can reach more people. Google first collaborated with the United States Geological Survey (USGS) and the California Governor's Office of Emergency Services to send these alerts to Android devices in California via the USGS’ ShakeAlert system. This uses information gathered from more than 700 seismometers across the state to figure out when an earthquake is taking place.... " 

Machine Learning and Earthquake Data

Been following this thread for sometime.  Been testing the CA use of local earthquake warning systems.   Actual predictive methods seem to be lagging here.  But the basic idea is very good.  Here aimed at detecting geothermal resources.

Machine Learning Picks Out Hidden Vibrations From Earthquake Data
MIT News
Jennifer Chu

Massachusetts Institute of Technology (MIT) researchers trained a convolutional neural network (CNN) with machine learning to identify low-frequency seismic vibrations in earthquake data. The researchers trained the network using the Marmousi model, a two-dimensional geophysical simulation of geological seismic-wave propagation. When presented with high-frequency seismic waves produced from a new simulated earthquake, the CNN could mimic the physics of wave propagation and accurately calculate the quake's hidden low-frequency waves. MIT's Laurent Demanet said the goal is to be able to use those low-frequency waves to map the Earth’s internal structures "and be able to say, for instance, 'this is exactly what it looks like underneath Iceland, so now you know where to explore for geothermal sources.'"

California Launches Earthquake Early Warning System

What is particularly good here is that the means to test this will be obvious.   I get USGS reports on earthquakes via an App, can filter them to mag 4.5 and higher, minutes after they occur.  So comparison to predictions should be straightforward.  Look forward to ongoing analysis.

California Launches Earthquake Early Warning System
Calls it the best in the world    in Reuters  By Dan Whitcomb 

California has launched the first statewide earthquake warning system in the U.S., aimed at detecting seismic waves and notifying residents via a mobile phone app up to 20 seconds before tremors hit. The California Earthquake Early Warning System taps hundreds of sensors to detect P-waves, which travel through the interior of the Earth, and arrive prior to surface waves and at a higher frequency during a temblor. University of California, Berkeley seismologists and engineers designed the MyShake phone app, which will initially warn users of local quakes with a magnitude of 4.5 or higher. Alerts are based on the ShakeAlert computer program operated by the U.S. Geological Survey, which analyzes data from seismic networks across the state, estimates the preliminary magnitudes of tremors, and calculates which areas will feel them.  ... "

Here is more about the early warning mobile App. I have it installed.

Advances in AI Earthquake Prediction

We attended some early neural net applications meeting where this was proposed, and added some of our own thoughts.   Nice to see this is evolving.   Are there shaking patterns in the earth that reliably predict earthquakes?    Thinking it is likely yes, but enough for the prediction and likely magnitude and location of major events?   I think yes too

AI Helps Seismologists Predict Earthquakes  in Wired
Machine learning is bringing seismologists closer to an elusive goal: forecasting quakes well before they strike. .... 

Artificial Intelligence Takes On Earthquake Prediction in QuantaMag

After successfully predicting laboratory earthquakes, a team of geophysicists has applied a machine learning algorithm to quakes in the Pacific Northwest.

In May of last year, after a 13-month slumber, the ground beneath Washington’s Puget Sound rumbled to life. The quake began more than 20 miles below the Olympic mountains and, over the course of a few weeks, drifted northwest, reaching Canada’s Vancouver Island. It then briefly reversed course, migrating back across the U.S. border before going silent again. All told, the monthlong earthquake likely released enough energy to register as a magnitude 6. By the time it was done, the southern tip of Vancouver Island had been thrust a centimeter or so closer to the Pacific Ocean.

Because the quake was so spread out in time and space, however, it’s likely that no one felt it. These kinds of phantom earthquakes, which occur deeper underground than conventional, fast earthquakes, are known as “slow slips.” They occur roughly once a year in the Pacific Northwest, along a stretch of fault where the Juan de Fuca plate is slowly wedging itself beneath the North American plate. More than a dozen slow slips have been detected by the region’s sprawling network of seismic stations since 2003.  And for the past year and a half, these events have been the focus of a new effort at earthquake prediction by the geophysicist Paul Johnson.    ..... " 

Predicting Earthquake Shocks

An example of application we saw early on for neural pattern prediction, now coming to fruition.  Here in Nature.  Other apparently subtle, but important links to patterns? 

Artificial intelligence nails predictions of earthquake aftershocks

A neural-network analysis outperforms the method scientists typically use to work out where these tremors will strike.

A machine-learning study that analysed hundreds of thousands of earthquakes beat the standard method at predicting the location of aftershocks.

Scientists say that the work provides a fresh way of exploring how changes in ground stress, such as those that occur during a big earthquake, trigger the quakes that follow. It could also help researchers to develop new methods for assessing seismic risk.

“We’ve really just scratched the surface of what machine learning may be able to do for aftershock forecasting,” says Phoebe DeVries, a seismologist at Harvard University in Cambridge, Massachusetts. She and her colleagues report their findings1 on 29 August in Nature.

Aftershocks occur after the main earthquake, and they can be just as damaging — or more so — than the initial shock. A magnitude-7.1 earthquake near Christchurch, New Zealand, in September 2010 didn’t kill anyone: but a magnitude-6.3 aftershock, which followed more than 5 months later and hit closer to the city centre, resulted in 185 deaths.  ... " 

Sensors for Earthquakes

We did a short test in this area:

Tiny Sensors May Help Avert Earthquake Damage, Track Sonar Danger, 'Listen' to Pipelines
Simon Fraser University
Marianne Meadahl

Engineers at Simon Fraser University (SFU) in Canada have developed ultra-sensitive accelerometers that are capable of capturing the most minuscule seismic activities. The new devices measure how tiny "seismic mass" comprised of silicon is displaced, due to external vibrations. The sensor can measure seismic mass displacements in the order of 1/10,000th of the diameter of a hydrogen atom. "The sensitivity of these devices is such that they can pick up the pressure waves produced by an earthquake before it strikes," says SFU's Behraad Bahreyni. The SFU team started out developing the accelerometers as a solution to make underwater sonar systems more compact and cost-efficient. However, they quickly realized that the high-performance, micro-machined accelerometers could have additional applications in detecting sound waves. ... " 

Predicting Earthquakes

Had seen a number of research efforts here,  another move ...

Today, A.I. helps detect tiny earthquakes. Tomorrow, it might predict the big one

Earthquakes are notoriously difficult to predict. Even major quakes often occur with little warning. Meanwhile, there are many hundreds of thousands of smaller earthquakes that humans rarely ever feel but are occasionally detected on seismographs.

Now, researchers from Harvard University and the Massachusetts Institute of Technology have developed an artificial intelligence (A.I.) neural network to better help detect earthquakes of all sizes. In a recent study published in the journal Science Advances, the A.I. system was shown to be more accurate than current methods, and may help bring seismologists closer to the elusive goal of earthquake prediction. .... " 

IoT Earthquake Warning Network

Warning versus predicting.

Zizmos Continues Its Quest to Create an IoT Earthquake-Warning Network    By Tekla S. Perry

 A simulation of Zizmos' earthquake early warning system shows the progression of a temblor in the San Francisco Bay Area

A few smartphone users in the Mexico City area were running the Zizmos app, described below, when this week’s magnitude-7.1 earthquake struck, Zizmos founder Battalgazi Yildirim reports, but not enough to issue a warning, although Zizmos did register the shaking.

Yildirim says he’d like to be able to get 50 fixed sensors installed in Mexico City—enough to reliably give warnings of aftershocks. The design, however, is still at the prototype stage, so each costs about $500 to build. He only has 10 on hand to donate, and would need funding to produce 40 more and local volunteers to install them.

Meanwhile, since the Mexico City earthquake, he says, another 5,000 smartphone users around the world have started running the app.

I first met Battalgazi Yildirim two years ago. He had posted a request in my local online community: His startup, Zizmos, wanted volunteers willing to mount a sensor package inside their homes, preferably on a bearing wall, to test whether a network of cheap packages of electronics, based on the Android phone design and his algorithms, could give early warnings of earthquakes. He wasn’t looking to do long-term prediction, just 15 or 30 seconds—enough to allow people to grab their kids and move to the safest spot in their house. .... " 

Predicting Earthquakes with Machine Learning

Broad idea has been around for years.  Are we getting closer?  Gets heavily into accuracy and risk of different kinds of errors here.    We worked with them on machine failure analysis and prediction, which has some similar profiles of failure.

Machine-learning earthquake prediction in lab shows promise  LANL press release. 

A computer science approach using machine learning can predict the time remaining before the fault fails   Listening to faultline’s grumbling gives countdown to future quakes

LOS ALAMOS, N.M., Aug. 30, 2017—By listening to the acoustic signal emitted by a laboratory-created earthquake, a computer science approach using machine learning can predict the time remaining before the fault fails.

“At any given instant, the noise coming from the lab fault zone provides quantitative information on when the fault will slip,” said Paul Johnson, a Los Alamos National Laboratory fellow and lead investigator on the research, which was published today in Geophysical Research Letters.

“The novelty of our work is the use of machine learning to discover and understand new physics of failure, through examination of the recorded auditory signal from the experimental setup. I think the future of earthquake physics will rely heavily on machine learning to process massive amounts of raw seismic data. Our work represents an important step in this direction,” he said.

Not only does the work have potential significance to earthquake forecasting, Johnson said, but the approach is far-reaching, applicable to potentially all failure scenarios including nondestructive testing of industrial materials brittle failure of all kinds, avalanches and other events.

Machine learning is an artificial intelligence approach to allowing the computer to learn from new data, updating its own results to reflect the implications of new information.

The machine learning technique used in this project also identifies new signals, previously thought to be low-amplitude noise, that provide forecasting inform with slow earthquakes on tectonic faults in the lower crust,” Johnson said. “There is reason to expect such signals from Earth faults in the seismogenic zone for slowly slipping faults.” ....  

Machine learning algorithms can predict failure times of laboratory quakes with remarkable accuracy. The acoustic emission (AE) signal, which characterizes the instantaneous physical state of the system, reliably predicts failure far into the future. This is a surprise, Johnson pointed out, as all prior work had assumed that only the catalog of large events is relevant, and that small fluctuations in the AE signal could be neglected.  .... " 

Predicting Earthquakes

A long time idea.   Will it work in the real world?

Intelligent Machines

Machine-Learning Algorithm Predicts Laboratory Earthquakes
The breakthrough has astonished geologists and raises the possibility that real earthquake prediction could be next.

by Emerging Technology from the arXiv  March 3, 2017  .... 

Crowdsourcing Earthquake Sensor Data

Prediction research via sensor data.   An update on the use of captured data.

Berkeley News

Quake-detection app captured nearly 400 temblors worldwide
By Robert Sanders, Media relations 

The Android app harnesses a smartphone’s motion detectors to measure earthquake ground motion, then sends that data back to the Berkeley Seismological Laboratory for analysis. The eventual goal is to send early-warning alerts to users a bit farther from ground zero, giving them seconds to a minute of warning that the ground will start shaking. That’s enough time to take cover or switch off equipment that might be damaged in a quake.

To date, nearly 220,000 people have downloaded the app, and at any one time, between 8,000 and 10,000 phones are active — turned on, lying on a horizontal surface and connected to a wi-fi network — and thus primed to respond.

An updated version of the MyShake app will be available for download today (Dec. 14) from the Google Play Store, providing an option for push notifications of recent quakes within a distance determined by the user, and the option of turning the app off until the phone is plugged in, which could extend the life of a single charge in older phones. .... " 

Update on Hyperloop

Update:  High speed, maglev, in-vacuum transportation of people and cargo in relatively narrow pipes. People stacked in a recumbent style of seat.  Windowless, but perhaps with virtual reality displays.

Impressive and futuristic  effort, probably doable in the longer term. Though I wonder about drilling through mountains, earthquakes, security, maintaining near vacuums and more.

This is also most useful at scale for long trips, where the infrastructure needs will be enormous.  And in  the short term, the slower, but very adaptable, competition from self driving vehicles.  I do like the idea of adding cargo potential. Unlike the Mars trip, here I would readily take a test trip.  Invite me. Following.

Bouncing Camera Gathers Images, Realities

Fast image gathering for reconnaissance and image based reality creation.  Possibilities.

Police Are Now Testing This Harvard Startup's Throwable Cameras
Bounce Imaging believes its technology could also have potential for a consumer action camera—as well as virtual reality.

Back in 2012, Bounce Imaging unveiled a prototype for the Explorer: a softball-sized throwable camera that could almost instantly transmit 360-degree images to a mobile device when tossed into a hidden areas, like a hostage situation, burning building or the rubble of an earthquake. The idea was originally conceived with first responders’ safety in mind, but it quickly became clear that there was immense value for giving police personnel an idea of what they may face in potentially dangerous situations. Now, three years later, the Boston startup is in the process of shipping out the first batch of 100 Explorers, which will be tested out by the Revere/North Metro SWAT team and the Maine Department of Corrections, among other departments.  ... "