This week’s list of data news highlights covers July 4-July 10, 2020, and includes articles about analyzing data to determine COVID-19 risk factors and a robot lab assistant that can perform some experiments 1,000 times faster than a human.
Researchers from multiple groups in the United Kingdom have analyzed the health records of more than 17 million English patients to determine factors that put individuals at higher risk of dying from COVID-19. The data, which comprises roughly 40 percent of England’s population, suggests that age is the most significant factor. Other variables that put a person at higher risk include if an individual is male, has underlying health conditions, or has a lower income.
Researchers from the University of Pittsburgh, University of Massachusetts Amherst, and Microsoft have developed an AI-enabled system that can identify the least energy-efficient buildings in an area. The system identifies inefficient buildings by comparing them to similar buildings in the same city or buildings in regions with similar weather. The researchers used the system to analyze data about buildings’ energy characteristics in three cities, finding that more than half of the residential buildings had inefficiencies.
Researchers from the University of Liverpool have developed a robot lab assistant that can perform some experiments 1,000 times faster than humans. Researchers can program the robot, which uses LiDAR to navigate, with basic experiment parameters. The robot then uses algorithms to decide how to change variables, such as the ratio of different chemical agents. For example, the robot performed nearly 700 experiments over eight days to determine substances that accelerate the process of creating hydrogen from light and water.
Researchers from the University of California, San Diego, have used a supercomputer to reveal the atomic makeup of the sugar-like molecules that help the coronavirus infect cells in the human body. The supercomputer performed simulations that showed these molecules change the shape of coronavirus proteins, which makes it easier for coronavirus to infect cells.
Researchers from MIT and Ava Robotics, a firm based in Massachusetts that develops robots, have created an autonomous robot that can disinfect a 4,000 square foot warehouse in 30 minutes. The groups built the robot, which uses UV-C light to clean surfaces, for the Greater Boston Food Bank. The researchers first teleoperated the robot to teach it how to navigate the food bank’s warehouse.
The state of Victoria, Australia, has sequenced roughly 80 percent of its COVID-19 cases, helping it identify specific outbreaks and infection-control failures. Researchers sequence the genome of the virus in different patients to identify linked transmissions, and researchers in Victoria used genomic sequencing to discover many early infections in Melbourne likely came from a “super spreader” of COVID-19. Sequencing also helped the state learn that a hotel quarantine program in May and June was not stopping the spread of coronavirus.
Transportation and logistics firm XPO Logistics is using an AI-powered robot to train employees on the layouts of its warehouses. The robot, which uses sensors such as laser pulses to determine the distance between itself and other objects, creates virtual maps of warehouses. It then teaches employees where items are in the warehouse.
Transport Genie, a technology firm based in Canada, has created a system of sensors to ensure the health and safety of livestock in transport. The system includes wireless sensors inside trailers that monitor factors such as temperature, humidity, and the braking and acceleration of a vehicle. The system can provide data to drivers in real-time and provide inputs to heating, cooling, and misting systems that can automatically activate at specific thresholds.
Microsoft has used AI to enable participants video chatting in Teams to be in a fixed virtual setting, such as in an auditorium or meeting room. The AI-enabled software creates a cutout of users’ live images and then places them in a shared environment to create a feeling that participants are in the same room.
Researchers from Ben-Gurion University of the Negev in Israel have developed an AI-enabled system that can predict the location of pilots flying drones in restricted airspace. The system uses a neural network, which the researchers trained on footage of drones flown by pilots in known locations, to analyze flight patterns. These patterns, such as a pilot’s tendency to fly a drone in a path around a central point, helped the system detect the location of operators with 78 percent accuracy.