Friday June 21, 2019
Home Lead Story New Artificia...

New Artificial Intelligence System Helps Drones Land More Quickly

Deep neural networks are capable of automatic learning, which makes them ideally suited for repetitive tasks

0
//
The enterprise solutions major has integrated SAP CoPilot with the
"A tectonic shift is happening in AI. Nearly 85 per cent of enterprises globally will use AI in some form or the other by 2020.

Researchers have developed an Artificial Intelligence (AI)-based system that can help drones to land more safely and quickly, while gobbling up less power.

The system, dubbed the “Neural Lander,” is a learning-based controller that tracks the position and speed of the drone, and modifies its landing trajectory and rotor speed accordingly to achieve the smoothest possible landing.

“This project has the potential to help drones fly more smoothly and safely, especially in the presence of unpredictable wind gusts, and eat up less battery power as drones can land more quickly,” said Soon-Jo Chung, Professor at California Institute of Technology (Caltech) in the US.

Landing multi-rotor drones smoothly is difficult. Complex turbulence is created by the airflow from each rotor bouncing off the ground as the ground grows ever closer during a descent.

This turbulence is not well understood nor is it easy to compensate for, particularly for autonomous drones.

That is why takeoff and landing are often the two trickiest parts of a drone flight. Drones typically wobble and inch slowly toward a landing until power is finally cut, and they drop the remaining distance to the ground.

Microsoft has announced a unique partnership with China-based DJI -- the world's biggest drone company -- where DJI will create a new software development kit (SDK) for Windows 10 PCs.
Representational image. Pixabay

The new system developed by researchers at Caltech uses a deep neural network to help autonomous drones “learn” how to land more safely and quickly.

Deep neural networks (DNNs) are AI systems that are inspired by biological systems like the brain.

The “deep” part of the name refers to the fact that data inputs are churned through multiple layers, each of which processes incoming information in a different way to tease out increasingly complex details.

Deep neural networks are capable of automatic learning, which makes them ideally suited for repetitive tasks.

Also Read: Tech Giant Apple Releases System Updates for iPhones, MacBooks

To make sure that the drone flies smoothly under the guidance of the deep neural networks, the team employed a technique known as spectral normalisation, which smoothens out the neural net’s outputs so that it doesn’t make wildly varying predictions as inputs/conditions shift.

“With less error, the Neural Lander is capable of a speedier, smoother landing and of gliding smoothly over the ground surface,” said Yisong Yue, Assistant Professor of Computing and Mathematical Sciences at Caltech. (IANS)

Next Story

Google Street Powers Artificial Intelligence Tool to Supervise Road Infrastructure

The fully-automated system is based on AI-powered object detection to identify street signs in the freely available images

0
Artificial intelligence, road infrastructure
The fully-automated system is based on AI-powered object detection to identify street signs in the freely available images. Pixabay

By tapping into Google Street View images, researchers have developed a new programme using artificial intelligence (AI) to monitor the stop and give way (yield) signs needing replacement or repair. The fully-automated system is based on AI-powered object detection to identify street signs in the freely available images.

Published in the journal of Computers, Environment and Urban Systems, the study shows the system detects signs with near 96 per cent accuracy, identifies their type with near 98 per cent accuracy and can record their precise geo-location from the 2D images.

artificial intelligence, road infrastructure
File:Opel Astra Google Street View. Wikimedia Commons

“The proof-of-concept model was trained to see ‘stop’ and ‘give way’ (yield) signs, but could be trained to identify many other inputs and was easily scalable for use by local governments and traffic authorities,” said the study lead author Andrew Campbell from RMIT University in Australia.

ALSO READ: Youtube Now Let Users Try Virtual Make-Up While Watching Tutorial in Real Time

Municipal authorities spend a large amount of time and money monitoring and recording the geo-location of traffic infrastructure manually, a task which also exposes workers to unnecessary traffic risks. “By using free and open source tools, we’ve developed a fully automated system for doing that job, and doing it more accurately,” Campbell said. (IANS)