This week’s list of data news highlights covers June 20-26, 2020, and includes articles about using AI to document war crimes and linking a volcanic eruption to the collapse of the Roman empire.
Researchers led by an individual from the California Institute of Technology have developed an AI-enabled system that has detected over 22,000 earthquakes, most of which were unknown to humans. The earthquakes ranged in magnitude from 0.7 to 4.4, and the software created one of the most detailed descriptions to date of earthquake swarms. These are strings of usually low-intensity earthquakes, and the system logged their time of occurrence and location.
A group of organizations led by Swansea University in the UK is using machine learning to document war crimes in Yemen. The organizations trained a system to analyze images and videos to detect BLU-63, a type of illegal munition that sprays out smaller explosives upon impact. The system can identify the munition with over 90 percent accuracy.
Honeywell, a company known for developing control systems for buildings and aircrafts, claims to have built the world’s highest-performing quantum computer. Honeywell measured its computer’s capabilities using quantum volume, a metric IBM created that uses a computer’s number of qubits, error rate, and how long the system can make calculations before the qubits stop working.
Researchers led by an individual from the California Institute of Technology have combined gravitational wave data and data from a robotic camera to find evidence that merging black holes could lead to an explosion of light. Gravitational wave data from May 21, 2019, indicates that black holes were spiraling toward each other. The researchers analyzed sky surveys, finding that a flare of light developed and faded in the same area and time as the merger of the black holes.
GNS Healthcare, a data analytics firm based in Massachusetts, has created a platform that uses AI to simulate how patients with a blood cancer called multiple myeloma might respond to different treatments. The firm built the platform using a dataset with information on the gene expressions, protein measurements, and attempted therapies of more than 1,000 patients. Drugmakers can use the platform to determine which patients to recruit for clinical trials.
The U.S. Food and Drug Administration has approved a bedside device from Circadia Health, a startup based in the United Kingdom, that uses AI to monitor respiratory health. The device detects a person’s breathing and chest movement using radar and sound, and algorithms can analyze the data to identify the early signs of COVID-19, such as breathlessness.
7. Deciphering Coronavirus Research
Primer AI, a startup based in San Francisco, has developed a public dashboard that highlights the latest research trends, news coverage, and social media discussion about coronavirus research. The platform uses natural language processing to analyze the more than 27,000 papers researchers have published about coronavirus to discover trends and write summaries. Individuals can use the platform to learn which papers are generating the most discussion and can sort the papers by categories, such as “forecasting and modeling.”
8. Linking Gene Mutations to Living Longer
Researchers from the University of Utah and the University of Louisville have analyzed the genetic material of three generations of individuals from 41 families to determine that developing fewer mutations is linked to living longer. The researchers found that individuals in the bottom 25 percent for the number of mutations outlived individuals in the top 75 percent by five years.
Researchers led by an individual from the Desert Research Institute in Nevada have used an array of sensors to determine that a volcanic eruption in Alaska may have helped spur the collapse of the Roman Empire. The researchers analyzed ice cores that corresponded to 43 BCE, and sensors found evidence of sulfur and volcanic ash. This evidence matched samples from a large volcanic eruption in Alaska, which could have caused substantial rain and the temperature to drop substaintially in the Mediterranean. Such conditions match accounts from Roman writers who noted that famines were occurring at the time.
Fugaku, a Japanese supercomputer, is now the fastest in the world. Riken and Fujitsu developed the computer, which is 2.8 times as fast as the previous record-holder, the United States’ Summit. Researchers have been using Fugaku to research coronavirus and simulate its spread.