Engineers have developed a robotic system that can evolve and improve its performance.
A robot arm builds “babies” that get progressively better at moving without any human intervention.
The ultimate aim of the research project is to develop robots that adapt to their surroundings.
The work by teams in Cambridge and Zurich has been published in the journal PLOS One.
It seems like a plot from a science fiction film: a robot that builds other robots – each one better than the previous generation. But that is what researchers in Cambridge and Zurich have done.
But those concerned about machines taking over the world shouldn’t worry, at least not yet.
At this stage the “baby robots” consist of plastic cubes with a motor inside. These are put together by a “mother” robot arm which glues them together in different configurations.
Although the set up is simple the system itself is ingenious.
The mother robot assesses how far its babies are able to move, and with no human intervention, improves the design so that the next one it builds can move further.
The mother robot built ten generations of children. The final version moved twice the distance of the first before its power ran out.
According to Dr Fumiya Iida of Cambridge University, who led the research with colleagues at ETH University in Zurich, one aim is to gain new insights into how living things evolve.
“One of the big questions in biology is how intelligence came about – we’re using robotics to explore this mystery,” he told BBC News.
“We think of robots as performing repetitive tasks, and they’re typically designed for mass production instead of mass customisation, but we want to see robots that are capable of innovation and creativity.”
Another aim is to develop robots that can improve and adapt to new situations, according to Andre Rosendo – who also worked on the project.
Read more: Robots learn to evolve and improve
The Latest on: Robots learn to evolve and improve
via Google News
The Latest on: Robots learn to evolve and improve
- India’s industrial automation market to grow over 10%: Sami Atiya, ABBon March 24, 2020 at 1:21 am
Machine learning allows the system to remember challenges and continuously improve item recognition and identification of optimal picking vectors for example. In the future, machine-learning and AI ...
- A Massive Opportunity Exists To Build ‘Picks And Shovels’ For Machine Learningon March 23, 2020 at 5:04 pm
The reward for building the next generation of developer tools will be many billions of dollars of enterprise value.
- To Be A Great Leader, Think Like An Exploreron March 21, 2020 at 8:04 am
The good news is that with technology we have access to learn anything we want at any time, right at our fingertips. Leaders with a growth mindset view challenges as opportunities to learn and improve ...
- Google taps computer vision to improve robot manipulation performanceon March 20, 2020 at 4:06 pm
In a preprint paper, a Google and MIT team investigate whether pre-trained visual representations can be used to improve a robot’s object manipulation performance. They say that their proposed ...
- BADGR uses deep learning to plan out and traverse obstacle-free pathson March 17, 2020 at 11:22 pm
BADGR can learn about navigational affordances, improve as it gathers more data, and generalize to unseen environments. Moving forward, the team stated that BADGR's success poses some questions.
- All the companies from Y Combinator’s W20 Demo Day, Part III: Hardware, Robots, AI and Developer Toolson March 17, 2020 at 12:49 pm
These are the companies that are working on hardware, robotics, AI, machine learning or tools for developers ... Daedalus is building autonomous software to allow industrial robots to operate without ...
- A weed-removing robot and an app for Alzheimer’s patients – meet this year’s Imagine Cup finalists from Kenya and Tunisiaon March 11, 2020 at 5:00 pm
Featuring an AI-powered camera, the robot can differentiate between weeds and crops, ensuring that only detrimental plants are removed. The team plans to improve the design further ... Placed at the ...
- Top Trends to Watch Out For in 2020 - Identify Major Growth Opportunities That Can Contribute to Your Growth Strategy - ResearchAndMarkets.comon March 6, 2020 at 4:57 am
Technologies that are able to combine different sources of information, such as LiDAR and imagery, will evolve rapidly to meet the ... Key innovation areas include agile robots with swift movements ...
- The Top Trends for 2020on March 2, 2020 at 8:33 pm
The regulatory landscape is also expected to evolve ... agile robots with swift movements and humanoid robots with ability to mimic and react to human thought processes in order to learn and ...
- Robot uses AI to personalize teaching of autistic childrenon February 27, 2020 at 8:56 am
Researchers have developed a new personalized learning robot for autistic ... Research suggests that robots can help autistic children improve both their academic and social skills.
via Bing News