This week’s list of top data news highlights covers October 29, 2022 to November 4, 2022 and includes articles on expanding autonomous vehicle operations and buying eyeglasses with augmented reality.
Google has updated its flood prediction and wildfire monitoring tools. The team trained the AI system behind the new flood forecasting tool with data on different types of drainage basins, which enables Google to scale the tool to include other countries more easily. The new version can now predict floods in 18 new countries. The company has also added a machine learning model to its wildfire tracking tool to improve wildfire monitoring.
Researchers at the University of California, Irvine have used a supercomputer to model flooding in Los Angeles during an extreme storm. The team found that between 197,000 and 874,000 residents would face a foot of flooding during such an event, higher than the federal government’s estimate of 23,000 residents, due to the city’s aging infrastructure.
Researchers at the University of California, Santa Cruz and Global Fishing Watch, a U.S.-based nonprofit organization dedicated to transparency in the fishing industry, have created an AI system that can analyze location data from boats’ beacons to determine if they are trying to evade detection. Many large boats are required to use beacons to relay their location in real time; however, some intentionally turn their beacons off to evade detection while fishing in illegal waters. The team used information from cases of purposeful beacon deactivation and unintentional deactivation, such as weak or intermittent signals, to train the system.
Researchers at the Madrid Institute for Advanced Studies in Materials Science have created a sensor that can detect dangerous levels of heat before a fire starts. Many common materials ignite between 300 to 500 degrees Celsius. The team’s sensor can detect fire dangers at 250 degrees Celsius by identifying changes in materials’ conductivity, and issue a warning in less than a second.
An international team of researchers has created a new technique to monitor the health of bridges. Currently, engineers monitor bridges through visual inspections or sensor technology; however, these methods are often costly and time-consuming. The researchers instead drove over bridges with smartphones and found that two smartphones’ accelerometer data could provide as accurate of data as 240 sensors. The team estimated that this technique could extend bridges’ lifespans by up to 30 percent.
The University of Hawaii Manoa and Hawaii Department of Health have launched a dashboard tracking substance use and mental health disorders in the state. The dashboard will show data on drug overdose deaths, calls made to crisis phone lines, and emergency room discharges related to mental health disorders or substance use or misuse. Public health officials can use the dashboard to design and implement prevention strategies and care programs.
OpenAI, a U.S.-based AI company, has launched an API for its text-to-image system DALL-E. Developers with other companies can use the API to integrate the system into their products.
Waymo has expanded its ride-hailing service to Phoenix’s Sky Harbor International Airport for riders in its Trusted Testers program, a group of approved individuals given early access to Waymo’s services. The company is the first autonomous vehicle operator to offer rides to a metropolitan airport.
Researchers at Massachusetts General Hospital have created an AI system that can predict the likelihood a patient with early-stage melanoma will experience a recurrence after initial treatments. The team trained the system with characteristics of 1,720 early-stage melanomas and patients’ medical data. Doctors can use the system to determine whether patients with early-stage melanoma would benefit from more intensive treatments.
Amazon and Snapchat have partnered to sell eyeglasses on Snapchat’s platform. Shoppers can use several augmented reality lenses to try on glasses and purchase them via an Amazon Fashion link on the screen.