PLUS: The latest AI and tech news.
By Jennifer Conrad | 11.08.21 Like R2-D2 and BB-8 from Star Wars, follower robots—artificially intelligent machines trained to trail people or other machines—can be tasked with carrying tools around a warehouse or just-picked fruit to a packing station. Now, makers of these robots want to coordinate movement around the modern workplace, Khari Johnson reports. Follower robots have been under development since the late 1990s. Initial forms relied on following the location of a tag in a person's pocket, but advances in deep learning and computer vision allow AI to navigate by "seeing" the world through cameras and other sensors. Today's bots can help workers manage heavy loads. Burro robots, which look like an autonomous driving pallet on the body of a four-wheel ATV, move freely between the rows of California fruit orchards and carry hundreds of pounds. In Switzerland, Gita robots follow food-delivery workers fulfilling orders and then make curbside deliveries. Too Much of a Good Thing? Theoretically, dozens or hundreds of robots can follow each other and work together. But the practice raises questions about how many robot helpers is too many before they become intimidating or an unhealthy reflection of work power dynamics. Read about follower robots' potential to help human workers. | A half century ago, sodium looked like a promising way to power a house or a car. But lithium was lighter and easier to work with than sodium, and a battery industry grew around it instead, writes Gregory Barber. So it was surprising when China's CATL, one of the world's largest battery makers, said that starting in 2023, it will begin placing sodium cells alongside lithium ones inside the battery packs that power electric cars. Along with CATL, a handful of labs and startups are giving sodium a second look. It's cheaper than lithium and performs better in cold weather. It's also well suited for applications such as managing the load within microgrids and keeping lights on at night when rooftop solar panels are in the dark. By the end of this decade, the world could be running short on the raw materials for batteries—not just lithium, but also other metals like nickel and cobalt. Sodium, on the other hand, can be found basically anywhere, including in seawater and in peat from bogs. New Challenges Sodium requires developing novel battery materials to avoid unwanted chemical reactions. It also packs less energy than lithium, meaning it can't power cars over long distances. Read about the work to develop sodium batteries. | Arctic air and water has warmed three times faster than the rest of the planet since 1971, which is causing sea ice to expand and contract in unpredictable ways. As Eric Niiler reports, scientists are now looking to artificial intelligence to aid in understanding what's happening at the ocean's surface. Information about the movement and location of ice cover is becoming increasingly valuable for tribal members in the Arctic, commercial fishers in places like Alaska, and global shipping companies interested in taking shortcuts through open patches of water. A model called IceNet provides a six-month forecast in each 25-kilometer-square grid across the region. The model is trained using deep learning, based on simulations of the Arctic climate in the years 1850 to 2100 and actual observational data recorded from 1979 to 2011. Tip of the Iceberg AI requires complex data and a lot of initial computing power, but once an algorithm is trained, it may be able to detect patterns in climate conditions thousands of times faster than physics-based models—and the algorithm gets better with time. Read how machine learning can help predict ice cover. | |
No comments:
Post a Comment