Connect with us


Robot beats humans at curling thanks to deep learning program



Sept. 23 (UPI) — Thanks to a new deep learning program, a curling robot, appropriately named Curly, was able to win three out of four matches against curlers from South Korea’s national teams.

Researchers detailed the software behind the breakthrough in a new paper published Wednesday in the journal Science Robotics.

Curling is sometimes described as a hybrid of bowling and chess — on ice. During gameplay, teams of two take turns throwing large “stones” across 150 feet of ice toward a bulls-eye target. The sport requires a combination of precise physical performance and strategic thinking.

“The game of curling can be considered a good testbed for studying the interaction between artificial intelligence systems and the real world,” Seong-Whan Lee, professor of cognitive engineering at Korea University, told UPI in an email.

Often, artificial intelligence system perform well in simulations, but struggle when applied in the real world. The problem is known as the “sim-to-real” gap.

In the computer lab, deep learning systems can learn from millions of actions in repeated simulations.

“In the real world, we may not even be able to perform hundreds of actions for the purpose of learning in each case,” Lee said. “Moreover, the system can never fully replicate the real world.”

Ice is one of many environments that is especially difficult to simulate. And because Curly must physically interact with the environment, simulation is even more difficult.

“In robotics, the sim-to-real gap has to do with visual perception, which means that the simulated world looks different from the real world,” Johannes Andreas Stork told UPI in an email.

Stork was not involved in the research, but wrote a commentary article on the research for Science Robotics.

“An example would be that a driverless car would see many more different cars and houses than one could possibly simulate,” said Stork, a professor of machine learning at Örebro University in Sweden.

In curling, with each throw, the ice conditions change. To compete against humans, researchers had to train Curly to judge uncontrollable environmental conditions and adapt.

Researchers supplied Curly with what they call a “deep reinforcement learning” system, a trial-and-error learning system that helps Curly compensate for uncertainties and take adaptive actions. Curly learns from each throw, allowing the robot to make corrections on subsequent throws.

“It is no longer necessary to identify the exact conditions on the ice sheet explicitly and therefore it is not necessary to a simulation that is exactly like the real world,” Stork said. “The simulation only had to change during training such that the policy has to learn to adapt. This is how the sim-to-real gap is addressed in this work.”

When the research team, scientists and engineers from Germany and Korea, combined their deep reinforcement learning system with a separate, previously developed strategy planning model, their artificial intelligence curling robot system was able to outperform expert curlers.

“We succeeded not only in terms of strategic planning but also with respect to the real-time adaptation within the real curling game setting,” Lee said.

Unlike human curling teams, which have three members, Curly uses only two robots — no sweepers. Curly relies on a skipper, the component in charge of aiming strategy, and a thrower, the component in charge of throwing mechanics.

Using its novel artificial intelligence system, the two components communicate to identify throwing errors, interpret changing ice conditions and make adjustments accordingly, all while accounting for the shifting strategy — dictated by the stones thrown by the human team.

Scientists hope their new deep reinforcement learning system can be adapted for a variety of complex real-world applications, including drone navigation.

“The approach presented in this work is suitable for problems where we have a task or environment that is changing over time and where the exact conditions are difficult to perceive from sensor data,” Stork said. “An industry-related scenario that I currently work with would be a plant for processing ore from a mine.”

“The composition of the ore changes over time, depending from where it was mined and the plant’s operation has to adapt to these changes,” he said. “Here, it is too expensive to analyze in a chemical lab on a continual basis.”

Source link

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *


NASA’s OSIRIS-REx touches down on asteroid Bennu to nab sample



Oct. 20 (UPI) — NASA’s OSIRIS-REx touched down on asteroid Bennu on Tuesday evening in a mission to scoop a sample of rocks and dirt.

The spacecraft — the Origins, Spectral Interpretation, Resource Identification, Security, Regolith Explorer — made soft contact with the asteroid at 6:12 p.m. EDT.

The historic “touch and go” event featured animation displaying OSIRIS-REx’s sample collection activities in real time. It takes time for real images of the touchdown to travel back to the Earth, so they won’t be released to the public until Wednesday.

The craft executed a series of maneuvers over the course of several hours before making soft contact with the surface of the asteroid to collect regolith, or rocks and dirt.

“It will be four and a half hours of anxiousness,” Beth Buck, OSIRIS-REx mission operations manager at Lockheed Martin Space, said in a news conference ahead of the event.

Buck made a comparison to the descent of a spacecraft on Mars, when there is typically “seven minutes of terror.”

The goal is to learn more about the solar system’s history and help “planetary defense” engineers with missions to protect earth from rogue asteroids. Bennu is believed to be a window into the solar system’s past since it’s a pristine, carbon-rich body carrying building blocks of both planets and life.

At around 1:50 p.m. EDT, the spacecraft left orbit around the asteroid before executing a series of burns to position itself over a sampling area nicknamed Nightingale.

Once in position, the craft began its approach to the asteroid at 5:50 p.m. EDT. It then spent about 15 seconds attempting to collect the regolith sample before backing away again.

The area, which is 52 feet in diameter, will make for a more demanding landing than expected, Kenneth Getzandanner, OSIRIS-REx flight dynamics manager at the NASA Goddard Space Flight Center in Maryland, said in the news conference.

The original mission called for a landing “zone” about 150% larger than Nightingale, at 82 feet, but that changed because Bennu was more rocky than expected.

The goal was to collect at least 1.7 ounces of fine-grained material, but the spacecraft can carry up to 4.4 pounds, Heather Enos, OSIRIS-REx deputy principal investigator at the University of Arizona said.

“I would love for that capsule to be completely full,” Enos said.

Though early images from the asteroid should hint at whether the mission succeeded, it will take engineers roughly 10 days to compare and analyze the mass before and after the maneuver to actually know how much dirt is inside the OSIRIS-REx.

If it failed, the spacecraft has enough fuel to attempt two more touch downs to collect material.

The spacecraft is expected to return to Earth, with the regolith sample from Bennu, in 2023.

Source link

Continue Reading


SpaceX scrubs Starlink launch until Thursday, if weather cooperates



Oct. 21 (UPI) — Just three days after sending 60 more Starlink satellites into orbit, SpaceX is aiming to launch another batch of broadband satellites into space from Florida.

If the weather cooperates, Thursday’s launch will be SpaceX’s 15th Starlink mission.

Liftoff had been scheduled for 12:29 p.m. EDT Wednesday aboard a Falcon 9 rocket at Launch Complex 40 at Cape Canaveral Air Force Station, but controllers scrubbed the launch due to weather and rescheduled for 12:14 p.m. on Thursday.

With a launch Sunday, SpaceX increased the size of their Starlink constellation to nearly 800 satellites. The 15th mission will see another 60-odd satellites join the network.

“The goal of Starlink is to create a network that will help provide Internet services to those who are not yet connected, and to provide reliable and affordable Internet across the globe,” according to the Kennedy Space Center.

Weather for Wednesday’s planned launch had looked so-so and the Space Force’s 45th Weather Squadron predicted a 60 percent chance of favorable conditions.

“A mid-level inverted trough and associated easterly wave currently across the Bahamas will meander into the state over the next few days, bringing enhanced moisture, cloud cover, and instability with a higher coverage of showers and storms,” Space Force forecasters wrote.

They said Thursday’s forecast looks quite similar to Wednesday’s.

Earlier this month, SpaceX founder and CEO Elon Musk tweeted that Starlink’s constellation was big enough to begin beta-testing the Internet service system in both the United States and southern Canada.

SpaceX has already offered Starlink Internet services to emergency responders in wildfire-stricken areas of Washington State.

Washington’s Hoh tribe is also using the Internet service to provide their members online education and telehealth services.

Source link

Continue Reading


Chernobyl-level radiation harms bumblebee reproduction



Oct. 21 (UPI) — Bees are more sensitive to radiation than scientists thought. Scientists found the reproduction rates of bumblebees declined significantly when exposed to Chernobyl-level radiation.

The research, published Wednesday in the journal Proceedings of the Royal Society B: Biological Sciences, suggests radiation in Ukraine’s Chernobyl Exclusion Zone could impair pollination services, triggering wider ecological consequences than previously estimated.

Humans are not allowed to live in the Chernobyl Exclusion Zone, the disaster area more directly impacted by the 1986 nuclear accident, the worst in history. However, the destroyed nuclear reactors are surrounded by forests that are populated by robust populations of birds, bears, bison, lynx, moose, wolves and more.

Efforts to gauge the effects of radiation contamination on insects have yielded mixed results in the past. While some studies have suggested insects are relatively radiation-resistant, others have demonstrated significant impairment.

When researchers exposed bumblebees in the lab to radiation dose of 100 µGyh-1, an amount approximating exposure inside the Chernobyl Exclusion Zone, reproduction rates among the bees dropped between 30 and 45 percent.

Researchers found a direct correlation between the size of the radiation dose and reproduction rate declines. Lower levels of radiation had a smaller effect, while larger doses yielded greater declines.

Scientists were surprised to find they were able to detect reproductive rate declines at very small levels of radiation exposure.

“Our research provides much needed understanding as to the effects of radiation in highly contaminated areas and this is the first research to underpin the international recommendation for the effects of radiation on bees,” lead study author Katherine Raines, environmental scientist at the University of Stirling in Scotland, said in a news release.

Source link

Continue Reading