This week’s list of data news highlights covers March 12, 2022 to March 18, 2022 and includes articles on finding meteorites with AI and using a supercomputer to determine the impact of climate change on El Niño episodes.
Researchers at McGill University in Canada have used an AI system to analyze the impact of psychedelic drugs on the brain. The team used AI to find the most commonly used words in 6,850 accounts of experiences taking one of 27 psychedelic drugs. They then combined the results with records of brain receptors known to be affected by each drug. Researchers can use the resultant patterns to identify potential drug treatments for mental health conditions.
Researchers at Northwestern University have created a microchip sensor with a structure that enables it to travel through the air by catching wind drafts. The sensor can collect data on air pollution and airborne diseases and transfer data to a smartphone or tablet with an onboard antenna.
Researchers at the University of Oxford have created an AI system that can model the size and vascularity of a placenta in real time and inform pregnant women about the potential for adverse conditions like preeclampsia. The team trained the system with thousands of ultrasound images of placentas. Doctors in the UK’s National Health Service launched a trial of the AI system this past week.
Researchers at Yale University have created a machine learning tool that can predict a patient’s mortality risks from COVID-19. The team created the tool by using an algorithm known as PHATE to analyze 55 million blood cells from 163 patients. They found that activity in certain immune cells was likely to indicate higher mortality risks. In tests, the tool predicted mortality risks with 83 percent accuracy.
Chipotle has launched a pilot project to use an autonomous robot to cook tortilla chips. The robot has an AI system that can recognize chip ingredients and manipulate a robotic arm into cooking them according to Chipotle’s recipe.
City officials in Houston, Texas have installed digital kiosks that can emit free Wi-Fi and provide information on municipal issues around the city. Each kiosk will display public safety notifications, real-time data on transit systems, a job board, and information about social services.
Researchers at the Massachusetts Institute of Technology have created an AI system that can optimize the performance of quadrupedal robots. Their quadruped robot named Mini Cheetah has an AI system that uses reinforcement learning, which enables the robot to learn through experimentation. In tests, the team achieved a new top speed for Mini Cheetah.
Researchers at the University of Texas at Austin, Rice University, and the University of Colorado, Boulder used a supercomputer to determine the impact of climate change on the severity of El Niños, a climate phenomenon that can cause adverse events like coral bleaching, where coral reef expel the algae that live in their tissue and die. They found that the frequency of severe El Niño episodes increased with time but that the impact of climatic shifts only had a small impact.
Researchers at Curtin University in Australia have created an AI system that can locate meteorites that have fallen to Earth using drone footage. The team trained the system with drone photos of known meteorites. Fallen meteorites typically take hundreds of days to find but, in April, the team used the system to find a fallen meteorite in just 3 days.
Researchers at the University of California, Berkeley and Google have created an AI system that can design AI chips. The team trained the system with blueprints of various AI chips. In tests against simulation-produced chips, the AI-created AI chips reduced latency by 2.7 times.