Home BlogWeekly News 10 Bits: the Data News Hotlist

10 Bits: the Data News Hotlist

by Joshua New
Nunavut ice sheet

This week’s list of data news highlights covers December 16, 2017 – January 5, 2018, and includes articles about a system of sensors that can make travelling across ice safer and an app that allows users to monetize their own health data.

1. Building an Ocean of Things

The U.S. Defense Advanced Research Projects Agency (DARPA) has launched a new program called the Ocean of Things to deploy thousands of floating sensors that can gather and transmit data about the environment and commercial vessel activity over large ocean areas. Monitoring large areas of the ocean is traditionally cost-prohibitive, but with low-cost commercial sensors, DARPA expects to be able to substantially expand its maritime awareness at a fraction of the cost of standard approaches. The sensors for the Ocean of Things project will be able to handle harsh ocean environments by using environmentally safe materials that can be disposed on the ocean floor at the end of their lifespan.

2. Making Hong Kong a Smart City

Hong Kong has published the “Smart City Blueprint,” a set of policies designed to promote the use of information technology to improve urban life, make Hong Kong more attractive for international businesses, and promote continuous innovation and sustainable economic development. For example, the Blueprint calls for installing 1,200 traffic sensors by 2020, deploying air sensors to monitor pollution, and providing all residents with electronic identification.

3. Keeping Kids Safe with Data Analytics

For the past 18 months, the Office of Children, Youth, and Families (C.Y.F.) in Allegheny County, Philadelphia has been using a predictive analytics system called the Allegheny Family Screening Tool to identify children with a high risk of being abused or neglected and help prioritize investigations. The system analyzes risk factors for abuse and neglect, such as a family history of drug and alcohol abuse, psychiatric problems, and prior C.Y.F. calls, to produce a risk score for a child that C.Y.F. employees can use to screen cases more efficiently than manual analysis and more easily identify high-risk cases warranting an intervention.

4. Keeping People Safe with Smart Ice

Researchers at the Memorial University of Newfoundland have successfully completed a trial of their SmartICE project, which uses sensors to measure sea ice thickness to identify safe routes, and will expand the project to five areas throughout Nunavut, where many Inuit communities need to travel across the ice to access food and fuel. The SmartICE program uses long poles with sensors on them driven into sea ice to measure ice thickness and snow cover, combines this data with data from sensors towed behind snowmobiles that travel popular routes, and generates color-coded maps of safe routes that people can access online.

5. Stopping Smoking with a Smartwatch

Israeli analytics company Somatix has developed a smartphone app called SmokeBeat that can analyze data from motion in a smartwatch to detect when wearers bring their hand to their mouth to smoke. SmokeBeat tracks these motions and notifies users about how much they smoke, how much money they spend on cigarettes, and how their habit changes over time, to prompt users to reduce the amount they smoke. Somatix says that tracking smoking alone with the app causes between a 15 and 20 percent reduction in smoking, and using the app’s prompts reduces smoking volume by an additional 40 percent.

6. Filtering Through Fluff in News with AI

Researchers at Google and the University of Pennsylvania have developed a machine learning system that can analyze written news articles and classify them based on how much actual news content they contain with 80 percent accuracy. The system scores a news article’s content density based on the presence of words more likely to convey content, such as “official” or “today,” and less likely to do so, such as “man” or “world,” and is effective across a wide range of topics, including sports, science, and business journalism.

7. Preventing the Yakuza from Getting Loans

Japan’s National Police Agency has granted Japanese banks access to its database about organized crime, or yakuza, activities so the bank can avoid issuing loans to gangsters. The system allows banks to identify whether a loan applicant is a member of the yakuza, affiliated with the yakuza but not an official member, or has no ties to the yakuza. It also allows police to provide confirmation in the event an applicant is a member or affiliate. Previously, banks had to rely on their own databases of yakuza activities sourced from police information and news reports.

8. Monetizing Your Own Health Data

A startup called CoverUS has developed a blockchain-based marketplace that allows users to upload and sell data from their electronic health records (EHRs) in exchange for cryptocurrency that they can exchange for discounts on participating insurance plans or use to pay for services that can improve their health, such as gym memberships. Users will be able to download a CoverUS app on their smartphone and populate it with their health data, which EHR operators are legally required to make available via an open application programming interface beginning in January 2018.

9. Putting Brainwaves Behind the Wheel

Nissan has developed a prototype “brain-to-vehicle,” or “B2V,” driving system that uses a sensor-laden skullcap that monitors a driver’s brainwaves to anticipate a driver’s actions, such as turning the wheel or accelerating, up to half a second before a driver makes them. Nissan designed the B2V system to eventually bridge the gap between manual and autonomous driving as self-driving car technology matures.

10. Using Machine Learning to Automate Data Science

Researchers at the Massachusetts Institute of Technology have developed a machine learning system that can automatically select and adjust data science modeling techniques for a user-specified task. Identifying appropriate modeling techniques for a particular task and fine-tuning a model’s parameters can be very labor intensive for data scientists. The researchers’ system can perform these tasks 100 times faster and produce equivalent, if not better, results, than humans.

Image: Mike Beauregard.

You may also like

Show Buttons
Hide Buttons