This week’s list of data news highlights covers December 31, 2016 – January 6, 2017 and includes articles about a new chatbot that can replace medical calls and a vineyard using the Internet of Things to protect against climate change.
England’s National Health Service (NHS) has partnered with startup Babylon Health to pilot an AI-powered chatbot as an alternative to NHS’s 111 telephone helpline, which provides medical advice for callers with non-life-threatening conditions. Instead of calling the helpline, users can write to the chatbot via a Babylon app to describe their symptoms and the chatbot will recommend a course of action, such as visiting a doctor, going to the emergency room, or taking an over-the-counter treatment. Like the 111 helpline, the chatbot is designed to reduce the number of unnecessary doctor’s visits.
Google’s Deepmind has revealed that its Go-playing AI software AlphaGo has been playing Go online anonymously against highly-ranked players around the world, including the highest ranked player Ke Jie, and has won 60 games without losing one. Under the username “Master,” Deepmind had AlphaGo play online as a form of unofficial testing of the software, which Deepmind has updated since it beat Go champion Lee Sedol in March 2016.
Jackson Family Wines in Sonoma County, California, has implemented an array of data-driven technologies to keep its vineyards operational in response to climate change, which is creating hotter and drier conditions that can reduce or disrupt grape harvests. The vineyards use sensors that monitor sap levels in grape vines and wirelessly sends this data to a computer that determines how much water different areas of the vineyards need. Additionally, drones equipped with sensors monitor vegetation color, which can indicate nutritional deficiencies or vines that don’t get as much water as they’re supposed to, which could indicate an irrigation leak. And since climate change can also cause more extreme colder weather, temperature sensors throughout the vineyard will automatically trigger wind machines that circulate warm air if it gets too cold.
Researchers at Carnegie Mellon University have developed an AI system named Libratus that will compete against human poker champions in Texas Hold’em in a tournament beginning January 11 for a $200,000 prize. Games such as Chess or Go are considered “perfect-information games,” meaning moves and opportunities are clearly defined, however poker is an “imperfect-information game” since one player can’t know what cards the other has, making it more challenging for AI. To address this, the researchers trained Libratus on 10160 possible plays and programed it to use game theory to identify the best strategy to play.
Carnival Cruise Line has introduced a new system of networked sensors for its cruise ships that will help Carnival better manage staff and offer new services for passengers. Passengers will be issues a small wearable sensor called an Ocean Medallion that can transmit data with other sensors throughout the ship, allowing staff to monitor, for example, when passengers leave their rooms so they can send cleaning staff, or when a large number of passengers head to the ship’s restaurants, which can prompt them to alert kitchen staff about increased demand. Additionally, passengers can use the Ocean Medallion to unlock their room doors, as a mobile wallet aboard the ship, and to help family members find each other aboard the ship.
The Pierce County, Washington, emergency management department and the U.S. Geological Survey (USGS) are implementing a substantial upgrade to the network of sensors and other technologies that monitor Mount Rainier to improve early warning efforts in the event of a disaster. Rapidly melting snow, landslides, and eruptions can trigger massive volcanic mudflows, called lahars, that can be deadly, such as Mount Rainier’s Oso lahar in 2014 that killed 43 people. Pierce County officials and USGS are upgrading the initial network of detection sensors installed in the 1990s to digital sensors that can collect more accurate data and more easily share it with emergency officials. Officials are also planning on expanding the area monitored by these sensors and doubling the number of early warning alarms that the sensors can trigger when they detect a lahar.
Gandhinagar, India, claims to be the first smart city in India after installing a variety of connected technologies and services throughout the city, including sensorized street lights, connected cameras that monitor traffic, digital signage, environmental sensors, and near-ubiquitous Wi-Fi connectivity. Gandhinagar was selected to participate in India’s Smart City Mission to develop 100 smart cities, and though the city did not make the shortlist for the first rounds of government funding, it was able to deploy these systems by partnering with telecom company Sterlite Technologies.
Startup CrowdAI has developed a new product that offers “AI-as-a-service” (AIAAS), allowing its customers to send in imagery from satellites, drones, or autonomous cars, and have objects in the images automatically classified by CrowdAI’s computer vision software. When a customer sends in their images, CrowdAI’s algorithms automatically classifies objects it identifies, such as farms, cars, or buildings in satellite imagery, and then human reviewers verify its accuracy before CrowdAI sends this data back to the customer.
Eight international cancer research organizations have published 19,000 de-identified genomic records on major cancers as part of the American Association for Cancer Research’s (AACR) Project Genomics Evidence Neoplasia Information Exchange (GENIE) initiative, which focus on spurring precision medicine research worldwide. The Project GENIE data includes genomic and clinical data on 59 types of cancers, including breast and lung cancer, and is one of the largest publicly available cancer genomic data sets. AACR will update the Project GENIE data on a quarterly basis.
Digital Jersey, a digitization initiative of the States of Jersey, a British Crown Dependency, has partnered with telecom company JT to build a long-range wide area network (LoRaWAN) to support the Internet of Things and make the States of Jersey a “smart island.” LoRaWANs are wireless networks designed to allow low-power devices—typically battery powered—communicate across long distances or in challenging environments that cannot support traditional telecom infrastructure.