This week’s list of data news highlights covers March 30-April 3, 2015 and includes articles about Facebook’s artificial intelligence systems and an effort to open source the human body’s data.
Democrata, a UK-based startup, is using predictive analytics to warn civil engineers and construction companies what is under the ground before they begin a new project. Democrata uses geospatial software originally developed for the British Geological Survey, which provides geoscientific data for the British government, that can predict where historical artifacts might be buried so construction companies can avoid digging into them, which can drive up costs and add long delays to the construction process.
Amazon has released a new Internet-connected product line called Dash Buttons to let consumers order household items with the press of a button. Dash Buttons are small, Internet-connected buttons synced with a smartphone and Amazon account that allow users to order home goods, like laundry detergent and toilet paper, with the press of a button. Dash Buttons are free, and 17 brands currently participate in the program.
The city of Boston is expanding its ParkBoston program, which allows drivers to pay for parking with their smartphones, to all of the city’s parking spaces after a successful trial period. ParkBoston supplements traditional coin-fed meters with a smartphone app that allows users to pay and add more time to meters remotely. ParkBoston will also collect data that can be used to better manage the city’s parking spaces.
Physicists at the University of Science and Technology in China have successfully demonstrated that a quantum computer can carry out machine learning tasks. The physicists used a technique that takes advantage of quantum entanglement, the phenomenon that occurs when two or more objects have such a strong relationship that measurement of one directly affects another. With this successful demonstration, should quantum computing technology progress as expected, such techniques could dramatically reduce the time required to process data required for machine learning tasks, from years, in some cases, to seconds.
The National Aeronautics and Space Administration (NASA) has created an open database of research conducted aboard the International Space Station. NASA hopes public access to the database will spur new research, increase the number of scientists participating in space station research, and identify knowledge gaps in existing research. The database is free to access but NASA requires users to register so it can see how the data is being used and by whom.
Microsoft and the Federal University of Minas Gerais have partnered to create the Traffic Prediction Project with the goal of predicting traffic jams up to an hour before they occur. The project will analyze a wide variety of traffic data, including data from transportation departments, Microsoft Bing’s traffic maps, cameras on the road, and driver’s social networks, to identify patterns that might contribute to traffic jams. Microsoft has tested a limited version of this model in various cities with 80 percent accuracy, and it expects the new data sources will make its predictions more reliable.
A biomedical research team has developed an imaging technique that lets users zoom into microscopic detail of images of the human body with the same algorithms used by Google Maps. The algorithms stitch together data from microscopic images that can vary greatly in size and detail, which have traditionally presented challenges to researchers trying to assemble this data for analysis. This technique allows researchers to more rapidly analyze this data and get a more detailed view of the human body, which the research team hopes will lead to new medical treatments.
Australian cargo company RightShip is using predictive analytics to better assess the risks of its ships. RightShip called on data scientists to predict the risk factors of its ships based on a variety of historical data, including inspections, past incidents, terminal feedback, and operators, from 75,000 ships. This data-driven assessment found certain factors influences risk more than previously expected by the existing risk-assessment model, which was constructed based off of expert opinions rather than data analysis.
A startup called Conceivable has created an app to help women better understand their fertility and improve the likelihood of pregnancy. Conceivable analyzes user-inputted data on body temperature, menstrual cycle, and time of ovulation to create personalized plans for women trying to conceive, offering nutrition and lifestyle advice that could increase fertility. Users can monitor their progress via a monthly report and visual dashboard that tracks changes in inputted data over time.
A new platform called InHerSight uses crowdsourced reviews to rank companies based on how equitable they are as an employer. Users review company policies and practices that might contribute to a hostile or unaccommodating work environment for women, including the amount of maternity leave offered and how well women are represented in company leadership. InHerSight’s creator hopes that companies will use this data to create work environments that are more attractive to women.