AGL 40.21 Increased By ▲ 0.18 (0.45%)
AIRLINK 127.64 Decreased By ▼ -0.06 (-0.05%)
BOP 6.67 Increased By ▲ 0.06 (0.91%)
CNERGY 4.45 Decreased By ▼ -0.15 (-3.26%)
DCL 8.73 Decreased By ▼ -0.06 (-0.68%)
DFML 41.16 Decreased By ▼ -0.42 (-1.01%)
DGKC 86.11 Increased By ▲ 0.32 (0.37%)
FCCL 32.56 Increased By ▲ 0.07 (0.22%)
FFBL 64.38 Increased By ▲ 0.35 (0.55%)
FFL 11.61 Increased By ▲ 1.06 (10.05%)
HUBC 112.46 Increased By ▲ 1.69 (1.53%)
HUMNL 14.81 Decreased By ▼ -0.26 (-1.73%)
KEL 5.04 Increased By ▲ 0.16 (3.28%)
KOSM 7.36 Decreased By ▼ -0.09 (-1.21%)
MLCF 40.33 Decreased By ▼ -0.19 (-0.47%)
NBP 61.08 Increased By ▲ 0.03 (0.05%)
OGDC 194.18 Decreased By ▼ -0.69 (-0.35%)
PAEL 26.91 Decreased By ▼ -0.60 (-2.18%)
PIBTL 7.28 Decreased By ▼ -0.53 (-6.79%)
PPL 152.68 Increased By ▲ 0.15 (0.1%)
PRL 26.22 Decreased By ▼ -0.36 (-1.35%)
PTC 16.14 Decreased By ▼ -0.12 (-0.74%)
SEARL 85.70 Increased By ▲ 1.56 (1.85%)
TELE 7.67 Decreased By ▼ -0.29 (-3.64%)
TOMCL 36.47 Decreased By ▼ -0.13 (-0.36%)
TPLP 8.79 Increased By ▲ 0.13 (1.5%)
TREET 16.84 Decreased By ▼ -0.82 (-4.64%)
TRG 62.74 Increased By ▲ 4.12 (7.03%)
UNITY 28.20 Increased By ▲ 1.34 (4.99%)
WTL 1.34 Decreased By ▼ -0.04 (-2.9%)
BR100 10,086 Increased By 85.5 (0.85%)
BR30 31,170 Increased By 168.1 (0.54%)
KSE100 94,764 Increased By 571.8 (0.61%)
KSE30 29,410 Increased By 209 (0.72%)
Editorials

Robot made out of tree branches learns to move on its own

Scientists have evolved robots enough that they can jog and even backflip or teach themselves to maneuver on their
Published July 6, 2019 Updated July 10, 2019

Scientists have evolved robots enough that they can jog and even backflip or teach themselves to maneuver on their own. This time researchers have made a robot out of tree branches that learns to move by itself.

The robot was made as a part of collaboration between researchers at the University of Tokyo and an AI firm called Preferred Networks. The unique robot is able to move around despite of its size, shape, or the variety of branches it’s made from.

Video Courtesy: Azumi Maekawa

Before the robot was made, the team weighed the branches, scanned them in 3D, and then uploaded it to a deep learning network. With this data, along with details about the type of servos being used for connecting the branches and their range of motions, the team allowed an AI to simulate how a prototype robot might move through deep reinforcement learning, explained IEEE Spectrum.

The simulation shows movements that are said to produce the longest walks. After trying out millions of possible movements, the robot was finally able to get from point A to point B by slowly crawling itself.

However, the researchers did have to make sure that any learned behavior that can easily snap the branches was avoided as its parts aren’t that strong as compared to other materials such as carbon fiber.

Moreover, though the robot might not be as useful as the rest, the team believes that it’s very flexible as it can be made by something as readily available as tree branches and can be available in timely manner, instead of depending on specialized parts or components, as per Gizmodo.

Copyright Business Recorder, 2019

Comments

Comments are closed.