This week’s list of data news highlights covers December 7-December 13, 2019, and includes articles about using the Internet-of-Things (IoT) to improve city parks and using AI to spot corruption.
Researchers led by an individual from NASA have mapped the wind patterns of Mars’ upper atmosphere. Using data from a NASA spacecraft that has been orbiting Mars since 2013, the researchers found that there are pockets within the atmosphere that have extremely variable wind patterns. The researchers also found that Mars’ mountains, which are significantly bigger than mountains on Earth, affect winds as high as 170 miles in the air.
Las Vegas is adding infrared sensors to its parks to detect litter, graffiti, and people. The sensors can help the city identify issues, such as an individual entering a park after hours. In addition, the sensors can help the city understand the foot traffic parks receive, which can help the city prioritize its investments in popular parks.
Researchers from the University of California, Merced, and Nvidia have developed an AI model that can automatically choreograph dances for music. The researchers trained the model using 361,000 clips representing ballet, Zumba, and hip-hop dancing. This process helped the model organize basic movements into dances according to input music.
Researchers led by an individual from the U.S. Geological Survey have used a supercomputer to predict the Earth’s terrestrial ecosystems will hold ten percent less carbon dioxide in 2100. The Earth’s terrestrial ecosystems hold roughly 30 percent of carbon dioxide emissions, and the researchers used the supercomputer to create simulations that combined global climate, land use, and emission models to make predictions.
Researchers from the Commonwealth Scientific and Industrial Research Organisation, Australia’s federal science agency, have developed an AI model that can estimate the maximum lifespans of extinct species, including early humans. The researchers trained the model on the known genomes of 252 species, teaching the model to learn the areas of animal genomes that relate to lifespan. On average, the model predicted lifespans within four years of the known maximum lifespans for the species.
Strella Biotechnology, a startup based in Philadelphia, has developed a sensor that predicts when apples will ripen. Apples emit heightened levels of a gas called ethylene when they ripen, which can trigger surrounding apples to ripen. Strella Biotechnology’s sensor detects levels of ethylene and communicates when apples will ripen to companies that store produce, helping them reduce food waste.
Teledyne Webb Research, a U.S. technology firm, and the Woods Hole Oceanographic Institution, a research institution based in Massachusetts, have developed autonomous robots that explored Kolumbo, an extremely active underwater volcanic mountain in the South Aegean Sea. The robots developed their own navigation plans, worked in coordination with one another, and have numerous sensors, ranging from mass spectrometers to thermometers. The robots discovered a new hydrothermal vent that had microbial life, and NASA, which founded the mission, eventually hopes to deploy the robots to the subterranean oceans of Jupiter’s moons.
Researchers from the University of South California have developed a robotic arm that can brush people’s hair, which can help stroke survivors and other individuals that have trouble brushing. The robot uses a camera to create a 3D map of a person’s head, which it uses to plan its brush strokes. In addition, the robot uses a comb with sensors that can detect skin, ensuring the robot does not harm the individual it is brushing.
Anheuser-Busch InBev, the maker of Budweiser, has developed BrewRight, a tool that uses AI to identify potentially illicit transactions. The tool assigns a risk score for transactions based on several factors, such as a vendor’s request for a rush payment. These scores help InBev identify patterns of corruption in everyday transactions, instead of reviewing a random sample of transactions for illegal payments.
Researchers from the University of Cincinnati and the Cincinnati Children’s Hospital Medical Center have developed a neural network that can analyze MRI images to detect attention deficit hyperactivity disorder (ADHD). The model analyzes the images to detect breakdowns in the connections between different regions in the brain, which studies suggest cause ADHD. In addition to the MRI images, the model used other patient data to have an 82 percent probability of classifying individuals with ADHD accurately.