This week’s list of data news highlights covers August 1-August 7, 2020, and includes articles about predicting coffee demand during the pandemic and modeling the solar system’s shape.
Researchers from MIT have developed an AI system that can predict if it is more likely to make an accurate decision than a human. The system uses one machine learning model to make a decision, such as determining if text is hate speech, and another to predict if a human is more likely to make a better decision. The second model uses a human’s track record to improve its prediction of the human’s performance iteratively.
Researchers led by an individual from Boston University have used data from NASA space missions to simulate how different types of charged particles are likely interacting with our solar system’s shape. The simulations revealed that a mixture of cold and hot matter likely have caused the solar system to be shaped like a croissant rather than a comet, which researchers had previously believed.
Walters-Storyk Design Group, a firm based in New York that designs rooms to improve their acoustics, has created an AI-enabled system that can optimize music and performance spaces. The system can predict how changes to a room, such as the placement of damping modules, would affect its acoustics. The firm is using the system to design recording studios for Spotify.
Tennis tournaments are using Hawk-Eye Live, an automated system that can make calls in tennis, to reduce the need for line judges at matches due to coronavirus. The system analyzes videos from multiple cameras in real-time to track and create a 3D representation of the ball’s path within millimeters. The U.S. Open will use the system on all but two courts.
Researchers from the Austrian Academy of Sciences and the European Space Agency have created a system that allows for more accurate tracking of space debris during daylight. The system uses a camera, telescope, and specialized filter to measure sunlight reflections off space debris. This data allows the researchers to detect the location of an object within one meter.
Callaway, a sporting goods company based in California, is using data and a supercomputer to design golf clubs. The firm’s AI system not only analyzes how players’ shots vary using different club designs but also helps design club faces themselves. The supercomputer then allows the firm to design thousands of iterations of clubs in a couple of days.
Keurig Dr Pepper, which owns Dr Pepper, Snapple, and Canada Dry, is using WiFi-enabled coffee brewers to forecast demand more accurately. The firm refitted 10,000 machines with WiFi for consumers who opted into the exercise, allowing the company to learn how frequently individuals were using their coffee machines. This data helped the organization adjust its production and order raw materials to match recent surges in people drinking coffee.
The New York Mets are using an AI-enabled system to authenticate players and check their temperatures as they enter its baseball stadium. The system, which New York-based firm Alclear created, uses facial recognition to identify players. Los Angeles FC, a Major League Soccer club, also plans to use the system to check whether fans are wearing masks once fans return to its stadium.
Researchers led by an individual from Nagoya University in Japan have used data about the magnetic fields on the Sun’s surface to develop a model that can predict solar flares. The data helped the model identify key characteristics of regions of the Sun that could produce solar flares, such as twisting magnetic field lines and areas where positive and negative magnetic fields are next to one another. The model accurately predicted seven of nine large solar flares from the last solar cycle.
IBM and the Michael J. Fox Foundation have developed a machine learning model that can help clinicians determine how advanced Parkinson’s disease is in a patient. The model considers factors such as medication that can mask symptoms of the disease, such as tremors. This process can help clinicians assess the progress of the disease in relation to symptoms.