This week’s list of data news highlights covers May 29, 2021 – June 4, 2021 and includes articles about tracking fish with facial recognition and developing wearable technology for cows.
Nvidia, a U.S. company that develops graphics chips, has launched a new AI supercomputer at the National Energy Research Scientific Computer Center in California. The supercomputer, known as Perlmutter, is designed to be the fastest computer in the world for AI. Researchers will use the system to create the largest 3D map of the universe by assembling data from a cosmic camera that can capture as many as 5,000 galaxies in a single shot.
Zelp, an agri-tech start-up based in the United Kingdom, has developed a methane-capturing wearable mask for cows. The device uses a built-in sensor to measure a cow’s methane production. When the methane levels get too high, the mask redirects the gas to a methane-absorbing filter and converts it to carbon dioxide. The device can also relay information from its sensors about the health and feeding activities of the cows. During trials, the masks cut methane emissions by 53 percent.
Technology company Qualcomm Technologies and professional services firm Jacobs have developed “vehicle-to-everything” technology to help municipal governments with traffic management. Officials can install sensors on city vehicles to communicate with traffic lights, pedestrian crossings, and similar roadside technologies. City officials in Peachtree Corners, Georgia have started using the technology to improve roadside communication and monitoring.
Researchers from the Icahn School of Medicine at Mount Sinai, New York have developed a machine learning model that can predict the likelihood that a COVID-19 patient will need dialysis treatment. COVID-19 can cause severe and lasting harm to organs, including kidney damage. The researchers trained algorithms based on data from electronic health records about the first 12 hours of a patient’s hospital stay. The algorithms predict death or dialysis treatment risks for day one, day three, day five, and day seven of a patient’s hospital stay, which can help hospitals better allocate resources for dialysis.
Whooshh Innovations, a start-up based in Seattle that develops tools for fish wildlife management and hydropower energy, has developed a facial recognition scanner to recognize Asian carp in Emiquon, a wetland wildlife refuge in Illinois. The scanner, named FishL Recognition, is attached to a 30-foot chute and uses six cameras to capture 18 shots of the fish as they glide through the water passage. Using this information, the chute will redirect the fish depending on whether they are native fish.
Walmart has developed a new augmented reality app that can help employees better identify back room inventory and stock shelves quickly. To find a particular piece of inventory among many boxes on a shelf, employees can take a picture or video of the shelves on the app. The app scans the codes on the boxes to find the right one and when it does, it will add a digital identifier, such as a green checkmark, beside the right box on the screen. The app helps reduce inventory search time from two and a half minutes to just forty-two seconds.
IBM researchers have created a method that uses aerial imagery, lidar data, and computer vision analysis techniques to measure the amount of carbon stored in trees. The researchers first analyzed remote imagery of Manhattan trees to determine their species. They then combined this information with lidar data, which measures the height, width, and volume of two-dimensional imagery, from all five boroughs in New York City. The researchers used the resulting data to estimate that trees in Manhattan account for 52,000 tons of stored carbon.
Researchers at the London School of Hygiene and Tropical Medicine have developed an algorithm that can identify patients at risk for developing pancreatic cancer from blood or urine tests. The researchers used electronic health records from patients diagnosed with pancreatic cancer for symptoms and health statuses in the two year period before their diagnosis. Using this data, researchers trained an algorithm to differentiate between patients who developed pancreatic cancer versus those who did not. The algorithm successfully identified 72 percent of high-risk patients and 59 percent of low-risk patients.
Telecommunications company Orange has partnered with Polish blockchain company Smartkey to improve access controls to buildings and city services across Poland. Several city governments already use Orange’s IoT platform to remotely manage city services such as street lighting control. The partnership builds upon this technology by integrating Smartkey’s blockchain solution, which ensures that the data from the IoT sensors cannot be falsified after its creation. When combined, the two platforms will enable authorized city residents to gain quicker access to specific buildings and city and commercial services. Eighty cities will participate in the initial rollout.
England’s national soccer team has partnered with GPS provider STATSports to choose which players should play in the final game of the European soccer championship. The Football Association will use data such as player distance and speed measurements to fully understand players’ capabilities and provide more helpful coaching instructions.
Image Credits: Flickr user Rockin’Rita