This week’s list of data news highlights covers February 16-22, 2019, and includes articles about autonomous vehicles understanding traffic cops hand signals’ and an AI assistant that takes drive-through orders.
The UK government has announced that it will pay up to £115 million for 1,000 students to earn AI doctorate degrees at 16 of its universities. As part of the initiative, companies including DeepMind and Cisco have pledged to fund 200 additional students to earn master’s degrees in AI. The funding is part of the UK’s AI Sector Deal, which is an agreement between the government and the nation’s AI sector to boost the UK’s global position as a leader in AI technologies.
Researchers from several universities, including Cornell University, have sequenced the entire genome of the great white shark. The genome reveals that great white sharks have DNA sequences that create powerful blood-clotting agents and scaffolding proteins that help produce new flesh to heal wounds quickly. The genome also reveals that the sharks have DNA sequences dedicated to tumor suppression, potentially helping it fight cancer.
AirHelp, a travel technology company, has developed new AI bots to automate the process of helping airline passengers receive compensation for delayed, canceled, or overbooked flights. One of the AI tools automatically reviews claims to decide if they have merit, while another automatically requests missing information from customers if necessary. AirHelp uses another AI bot, called Lara, to gauge the likelihood a claim will have a successful outcome for customers with 96 percent accuracy.
Waymo has taught its autonomous vehicles to understand the hand signals of traffic cops. The cars use high-resolution cameras, radar, LIDAR, and AI software to detect the presence of officers and interpret their hand motions, allowing the vehicle to stop and continue driving based on the gestures of officers.
San Francisco is installing sensors in 1,000 trash bins to optimize waste collection. The sensors, which come from waste management company Nordsense, collect and relay data on the fullness, temperatures, and fill rates of the bins. Nordsense’s AI technology uses the data to find waste patterns and make predictions to help the city improve its waste collection. During a three-month pilot in 2018, the sensors reduced the number of overflowing bins by 80 percent.
A group of European and American researchers has used the data of more than 1,300 cancer patients undergoing chemotherapy to identify links between the treatment’s different symptoms. The researchers examined 38 different symptoms patients experienced, such as hot flashes, and analyzed their occurrence, severity, and resulting mental or physical anguish. Their research shows that some symptoms, such as nausea, are correlated with other symptoms, suggesting that alleviating such symptoms could simultaneously alleviate others.
DARPA is providing the Center for Open Science, a nonprofit organization that promotes open research, $7.6 million to create a database of 30,000 claims from social science research to help develop AI systems that assign credibility scores to claims. For 3,000 of the claims, the nonprofit will attempt to replicate them or ask humans to predict their replicability. In March, DARPA will start accepting applications for algorithms that can predict the replicability of research better than humans.
A Brazilian startup called Treevia has developed a remote monitoring system that uses sensors and machine learning to help companies and researchers track and predict the health of forests. The system’s sensors wrap around trees and capture changes in the diameter of the trees. Treevia combines this data with satellite images and climate data to predict how much forests will grow.
Good Times Burgers & Frozen Custard, a fast food restaurant in Denver, has implemented an AI voice assistant that takes orders at the drive-through window. The assistant reduced order time by roughly 25 percent during a testing phase that started in the fall of 2018. The assistant also frees employees to perform other tasks and alleviates them from having to say the same welcome language repeatedly throughout their shift.
Hospitals are increasingly using “smart” ID bracelets equipped with sensors and tags to track the health of their patients. For example, Wake Forest Baptist Health in Winston-Salem, North Carolina, uses such bracelets to ensure cancer patients exercise enough after surgery by tracking how often and for how long they are moving out of their rooms. In addition, St. Joseph Mercy Oakland Hospital in Pontiac, Michigan, uses the bracelets as part of a location system to know more precisely when patients leave the hospital, allowing them to clean beds and fill vacancies more efficiently.
Image: Terry Goss