Developed by UZH researchers, the algorithm DroNet allows drones to fly completely by themselves through the streets of a city and in indoor environments. Therefore, the algorithm had to learn traffic rules and adapt training examples from cyclists and car drivers.
All today’s commercial drones use GPS, which works fine above building roofs and in high altitudes. But what, when the drones have to navigate autonomously at low altitude among tall buildings or in the dense, unstructured city streets with cars, cyclists or pedestrians suddenly crossing their way? Until now, commercial drones are not able to quickly react to such unforeseen events.
Integrate autonomously navigating drones
Researchers of the University of Zurich and the National Centre of Competence in Research NCCR Robotics developed DroNet, an algorithm that can safely drive a drone through the streets of a city. Designed as a fast 8-layers residual network, it produces two outputs for each single input image: a steering angle to keep the drone navigating while avoiding obstacles, and a collision probability to let the drone recognise dangerous situations and promptly react to them. “DroNet recognises static and dynamic obstacles and can slow down to avoid crashing into them. With this algorithm we have taken a step forward towards integrating autonomously navigating drones into our everyday life”, says Davide Scaramuzza, Professor for Robotics and Perception at the University of Zurich.
Powerful artificial intelligence algorithm
Instead of relying on sophisticated sensors, the drone developed by Swiss researchers uses a normal camera like that of every smartphone, and a very powerful artificial intelligence algorithm to interpret the scene it observes and react accordingly. The algorithm consists of a so-called Deep Neural Network. “This is a computer algorithm that learns to solve complex tasks from a set of ‘training examples’ that show the drone how to do certain things and cope with some difficult situations, much like children learn from their parents or teachers”, says Prof. Scaramuzza.
Cars and bicycles are the drones’ teachers
One of the most difficult challenges in Deep Learning is to collect several thousand ‘training examples’. To gain enough data to train their algorithms, Prof. Scaramuzza and his team collected data from cars and bicycles, that were driving in urban environments. By imitating them, the drone automatically learned to respect the safety rules, such as “How to follow the street without crossing into the oncoming lane”, and “How to stop when obstacles like pedestrians, construction works, or other vehicles, block their ways”. Even more interestingly, the researchers showed that their drones learned to not only navigate through city streets, but also in completely different environments, where they were never taught to do so. Indeed, the drones learned to fly autonomously in indoor environments, such as parking lots and office’s corridors.
Toward fully autonomous drones
This research opens potential for monitoring and surveillance or parcel delivery in cluttered city streets as well as rescue operations in disaster urban areas. Nevertheless, the research team warns from exaggerated expectations of what lightweight, cheap drones can do. “Many technological issues must still be overcome before the most ambitious applications can become reality,” says PhD student Antonio Loquercio.
The Latest on: Automated drones
via Google News
The Latest on: Automated drones
- Russia’s Formidable Tunguska Air Defense System is Getting a Stealth Upgradeon November 30, 2019 at 4:10 am
will get passive thermal and optical sensors that will enable it to detect aircraft and drones without disclosing its position by emitting radar beams. “Thanks to the introduction of television and ...
- Can Amazon drones or Mars rovers be hacked? Very easily, UBC research suggestson November 29, 2019 at 3:04 pm
As automated drones and robotic vehicles become more popular — and are employed by everyone from Amazon to NASA — new B.C. research suggests the machines can be hacked with relative ease. Researchers ...
- IBM Patents Blockchain to Stop Drones From Stealing Packageson November 29, 2019 at 4:06 am
The patent seeks to get ahead of two modern realities: people buy goods online, and people fly their own personal drones. That could be a problem, it says, if the trends combine to devious ends.
- Cameras and drones: The best photographic gifts for shutterbugson November 28, 2019 at 5:00 am
If one of them is on your holiday gift list, check out CNET’s top picks to give to your favorite photographer, including three cameras and one very advanced drone. The good: The DJI Mavic Air’s ...
- UBC research highlights need to safeguard drones and robotic cars against cyber attackson November 27, 2019 at 6:44 am
Robotic vehicles like Amazon delivery drones or Mars rovers can be hacked more easily than people may think ... and it's these deviations that attackers can exploit to throw the vehicles off course.
- FAA Expands LAANC Drone Programon November 27, 2019 at 12:37 am
The FAA announced two important expansions of its automated drone flight application and airspace approval process for drones—the Low Altitude Authorization and Capability (LAANC) program. The agency ...
- Bowl Of Noodles Delivered Via Drone In Chinaon November 27, 2019 at 12:18 am
With a take-off weight slightly under 55 pounds, and a payload of up to 8.8 pounds, this drone flies at an altitude of less than 400 feet at up to 23 knots. The objective of the trial is to validate a ...
- US gridlock solutions envisioned with driverless cars, flying taxis and delivery droneson November 26, 2019 at 6:41 pm
While none of the 735 transportation technology leaders at the seventh annual Florida Automated Vehicles Summit described a world that will resemble ... a modular multi-story building where your ...
- First drone cargo delivery of an Airbus & XAG joint development in Guangzhou, China, codenamed "Project Vesper"on November 26, 2019 at 5:10 am
With a take-off weight slightly under 25 kilos, and a payload of up to 4 kilos, this drone flies at an altitude of less than 400 ft at up to 12 meters/second. The objective of the trial is to validate ...
via Bing News