A machine learning-based approach that could help astronomers detect lower-frequency gravitational waves has been unveiled by researchers in the UK, US, and Italy. Dubbed deep loop shaping, the system would apply real-time corrections to the mirrors used in gravitational wave interferometers. This would dramatically reduce noise in the system, and could lead to a new wave of discoveries of black hole and neutron star mergers – according to the team.

In 2015, the two LIGO interferometers made the very first observation of a gravitational wave: attributing its origin to a merger of two black holes that were roughly 1.3 billion light–years from Earth.

Since then numerous gravitational waves have been observed with frequencies ranging from 30–2000 Hz. These are believed to be from the mergers of small black holes and neutron stars.

So far, however, the lower reaches of the gravitational wave frequency spectrum (corresponding to much larger black holes) have gone largely unexplored. Being able to detect gravitational waves at 10–30 Hz would allow us to observe the mergers of intermediate-mass black holes at 100–100,000 solar masses. We could also measure the eccentricities of binary black hole orbits. However, these detections are not currently possible because of vibrational noise in the mirrors at the end of each interferometer arm.

Subatomic precision

“As gravitational waves pass through LIGO’s two 4-km arms, they warp the space between them, changing the distance between the mirrors at either end,” explains Rana Adhikari at Caltech, who is part of the team that has developed the machine-learning technique. “These tiny differences in length need to be measured to an accuracy of 10^-19 m, which is 1/10,000^th the size of a proton. [Vibrational] noise has limited LIGO for decades.”

To minimize noise today, these mirrors are suspended by a multi-stage pendulum system to suppress seismic disturbances. The mirrors are also polished and coated to eliminate surface imperfections almost entirely. On top of this, a feedback control system corrects for many of the remaining vibrations and imperfections in the mirrors.

Yet for lower-frequency gravitational waves, even this subatomic level of precision and correction is not enough. As a laser beam impacts a mirror, the mirror can absorb minute amounts of energy – creating tiny thermal distortions that complicate mirror alignment. In addition, radiation pressure from the laser, combined with seismic motions that are not fully eliminated by the pendulum system, can introduce unwanted vibrations in the mirror.

The team proposed that this problem could finally be addressed with the help of artificial intelligence (AI). “Deep loop shaping is a new AI method that helps us to design and improve control systems, with less need for deep expertise in control engineering,” describes Jonas Buchli at Google DeepMind, who led the research. “While this is helping us to improve control over high precision devices, it can also be applied to many different control problems.”

Deep reinforcement learning

The team’s approach is based on deep reinforcement learning, whereby a system tests small adjustments to its controls and adapts its strategy over time through a feedback system of rewards and penalties.

With deep loop shaping, the team introduced smarter feedback controls for the pendulum system suspending the interferometer’s mirrors. This system can adapt in real time to keep the mirrors aligned with minimal control noise – counteracting thermal distortions, seismic vibrations, and forces induced by radiation pressure.

“We tested our controllers repeatedly on the LIGO system in Livingston, Louisiana,” Buchli continues. “We found that they worked as well on hardware as in simulation, confirming that our controller keeps the observatory’s system stable over prolonged periods.”

Based on these promising results, the team is now hopeful that deep loop shaping could help to boost the cosmological reach of LIGO and other existing detectors, along with future generations of gravitational-wave interferometers.

“We are opening a new frequency band, and we might see a different universe much like the different electromagnetic bands like radio, light, and X-rays tell complementary stories about the universe,” says team member Jan Harms at the Gran Sasso Science Institute in Italy. “We would gain the ability to observe larger black holes, and to provide early warnings for neutron star mergers. This would allow us to tell other astronomers where to point their telescopes before the explosion occurs.”

The research is described in Science.

The post LIGO could observe intermediate-mass black holes using artificial intelligence appeared first on Physics World.

The global network of Android smartphones makes a useful earthquake early warning system, giving many users precious seconds to act before the shaking starts. These findings, which come from researchers at Android’s parent organization Google, are based on a three-year-long study involving millions of phones in 98 countries. According to the researchers, the network’s capabilities could be especially useful in areas that lack established early warning systems.

“By using Android smartphones, which make up 70% of smartphones worldwide, the Android Earthquake Alert (AEA) system can help provide life-saving warnings in many places around the globe,” says study co-leader Richard Allen, a visiting faculty researcher at Google who directs the Berkeley Seismological Laboratory at the University of California, Berkeley, US.

Traditional earthquake early warning systems use networks of seismic sensors expressly designed for this purpose. First implemented in Mexico and Japan, and now also deployed in Taiwan, South Korea, the US, Israel, Costa Rica and Canada, they rapidly detect earthquakes in areas close to the epicentre and issue warnings across the affected region. Even a few seconds of warning can be useful, Allen explains, because it enables people to take protective actions such as the “drop, cover and hold on” (DCHO) sequence recommended in most countries.

Building such seismic networks is expensive, and many earthquake-prone regions do not have them. What they do have, however, is smartphones. Most such devices contain built-in accelerometers, and as their popularity soared in the 2010s, seismic scientists began exploring ways of using them to detect earthquakes. “Although the accelerometers in these phones are less sensitive than the permanent instruments used in traditional seismic networks, they can still detect tremors during strong earthquakes,” Allen tells Physics World.

A smartphone-based warning system

By the late 2010s, several teams had developed smartphone apps that could sense earthquakes when they happen, with early examples including Mexico’s SkyAlert and Berkeley’s ShakeAlert. The latest study takes this work a step further. “By using the accelerometers in a network of smartphones like a seismic array, we are now able to provide warnings in some parts of the world where they didn’t exist before and are most needed,” Allen explains.

Working with study co-leader Marc Stogaitis, a principal software engineer at Android, Allen and colleagues tested the AEA system between 2021 and 2024. During this period, the app detected an average of 312 earthquakes a month, with magnitudes ranging from 1.9 to 7.8 (corresponding to events in Japan and Türkiye, respectively).

For earthquakes of magnitude 4.5 or higher, the system sent “TakeAction” alerts to users. These alerts are designed to draw users’ attention immediately and prompt them to take protective actions such as DCHO. The system sent alerts of this type on average 60 times per month during the study period, for an average of 18 million individual alerts per month. The system also delivered lesser “BeAware” alerts to regions expected to experience a shaking intensity of 3 or 4.

To assess how effective these alerts were, the researchers used Google Search to collect voluntary feedback via user surveys. Between 5 February 2023 and 30 April 2024, 1 555 006 people responded to a survey after receiving alerts generated from an AEA detection. Their responses indicated that 85% of them did indeed experience shaking, with 36% receiving the alert before the ground began to move, 28% during and 23% after.

Principles of operation

AEA works on the same principles of seismic wave propagation as traditional earthquake detection systems. When an Android smartphone is stationary, the system uses the output of its accelerometer to detect the type of sudden increase in acceleration that P and S waves in an earthquake would trigger. Once a phone detects such a pattern, it sends a message to Google servers with the acceleration information and an approximate location. The servers then search for candidate seismic sources that tally with this information.

“When a candidate earthquake source satisfies the observed data with a high enough confidence, an earthquake is declared and its magnitude, hypocentre and origin time are estimated based on the arrival time and amplitude of the P and S waves,” explains Stogaitis. “This detection capability is deployed as part of Google Play Services core system software, meaning it is on by default for most Android smartphones. As there are billions of Android phones around the world, this system provides an earthquake detection capability wherever there are people, in both wealthy and less-wealthy nations.”

In the future, Allen says that he and his colleagues hope to use the same information to generate other hazard-reducing tools. Maps of ground shaking, for example, could assist the emergency response after an earthquake.

For now, the researchers, who report their work in Science, are focused on improving the AEA system. “We are learning from earthquakes as they occur around the globe and the Android Earthquake Alerts system is helping to collect information about these natural disasters at a rapid rate,” says Allen. “We think that we can continue to improve both the quality of earthquake detections, and also improve on our strategies to deliver effective alerts.”

The post Android phone network makes an effective early warning system for earthquakes appeared first on Physics World.