Practise makes perfect – it is an adage that has helped humans become highly dexterous, and now it is an approach that is being applied to robots.
Computer scientists at the University are using the artificial intelligence (AI) techniques of automated planning and reinforcement learning to “train” a robot to find an object in a cluttered space, such as a warehouse shelf or in a fridge – and move it.
The aim is to develop robotic autonomy, so the machine can assess the unique circumstances presented in a task and find a solution – akin to a robot transferring skills and knowledge to a new problem.
The Leeds researchers are presenting their findings today at the International Conference on Intelligent Robotics and Systems in Macau, China. Their paper can be read here.
“ Robots aren’t very good at what humans do very well: being highly mobile and dexterous. Those physical skills have been hardwired into the human brain.”
The big challenge is that in a confined area, a robotic arm may not be able to grasp an object from above. Instead it has to plan a sequence of moves to reach the target object, perhaps by manipulating other items out of the way. The computer power needed to plan such a task is so great, the robot will often pause for several minutes. And when it does execute the move, it will often fail.
Developing the idea of practise makes perfect, the computer scientists at Leeds are bringing together two ideas from AI.
One is automated planning. The robot is able to “see” the problem through a vision system, in effect an image. Software in the robot’s operating system simulates the possible sequence of moves it could make to reach the target object.
But the simulations that have been “rehearsed” by the robot fail to capture the complexity of the real world and when they are implemented, the robot fails to execute the task, for example by knocking objects off a shelf.
The Leeds team have therefore combined planning with another AI technique called reinforcement learning.
Reinforcement learning involves the computer in a sequence of trial and error attempts – about 10,000 in total – to reach and move objects. Through these trial and error attempts, the robot “learns” which actions it has planned are more likely to end in success.
The computer undertakes the learning itself, starting off by randomly selecting a planned move that might work. But as the robot learns from trial and error, it becomes more adept at selecting those planned moves that have a greater chance of being successful.
Dr Matteo Leonetti, from the School of Computing, said: “Artificial intelligence is good at enabling robots to reason – for example, we have seen robots involved in games of chess with grandmasters.
“But robots aren’t very good at what humans do very well: being highly mobile and dexterous. Those physical skills have been hardwired into the human brain, the result of evolution and the way we practise and practise and practise.
“And that is an idea that we are applying to the next generation of robots.”
Wissam Bejjani, the PhD researcher who wrote the research paper, said the robot develops an ability to generalise, to apply what it has planned to a unique set of circumstances.
He added: “Our work is significant because it combines planning with reinforcement learning. A lot of research to try and develop this technology focuses on just one of those approaches.
“Our approach has been validated by results we have seen in the University’s robotics lab.
“With one problem, where the robot had to move a large apple, it first went to the left side of the apple to move away the clutter, before manipulating the apple.
“It did this without the clutter falling outside the boundary of the shelf.”
Dr Mehmet Dogar, Associate Professor in the School of Computing, was also involved in the study. He said the approach had speeded up the robot’s “thinking” time by a factor of ten – decisions that took 50 seconds now take five.
Learn more: Worker robots that learn from mistakes
The Latest on: Robotic autonomy
via Google News
The Latest on: Robotic autonomy
- Amazon chops up to $201 off these iRobot Roomba robot vacuumson January 9, 2020 at 6:03 am
It has three cleaning modes: One-cleaning pass, two-cleaning pass, and automatic. The first mode makes the Roomba 960 pass over an area once, the second twice, and the third gives the robot autonomy ...
- VW announces new Silicon Valley self-driving nerve center at CESon January 8, 2020 at 10:29 am
The newly created Volkswagen Autonomy, Inc., will be part of Volkswagen Autonomy GmbH, the umbrella organization within the German automaker charged with developing and commercializing robotic cars ...
- Toyota is building a tiny utopian prototype city filled with people, robots and AIon January 7, 2020 at 4:03 pm
“If they build it, they will come.” “Imagine a fully controlled site that will allow researchers, engineers and scientists the opportunity to freely test technology such as autonomy, ...
- Warehouse Robotics Startups Drawing Bigger Investor Backingon January 7, 2020 at 1:45 pm
Investors are pouring money into warehouse automation as more companies turn to robotics to help speed distribution in a tight labor market.
- Vecna Robotics raises $50 million to automate warehouse tasks with AIon January 7, 2020 at 8:05 am
... to develop a humanoid robot (the Battlefield Extraction-Assist Robot, or BEAR) capable of rescuing wounded soldiers from the battlefield. However, ahead of its incorporation in 2018 as a wholly ...
- Robotic Research unveils Pegasus Mini hybrid UAS/UGVon January 7, 2020 at 2:10 am
Robotic Research, a Maryland-based developer of autonomy and robotic technologies catering to commercial and government customers, has taken the opportunity at the CES 2020 exhibition being held in ...
- Robotic Research To Unveil Pegasus Mini At CES 2020on January 7, 2020 at 1:40 am
Robotic Research LLC, a leading provider of autonomy and robotic technologies to commercial and federal customers, will be debuting a new football-size transformable drone, the Pegasus Mini, at CES ...
- Robotics Industry to Shift from Fixed Automation to Mobile Systemson January 4, 2020 at 9:45 am
By 2022, the burgeoning mobile robotics space will start to overtake the traditional industrial robotics market. Currently, mobile autonomy is concentrated in material handling within the supply chain ...
- CES 2020: Robotic Research To Debut 1st Football-size Droneon January 4, 2020 at 5:35 am
Autonomy and robotic technology company Robotic Research will launch its first-ever football-size transformable drone at the upcoming Consumer Electronics Show from January 7 to 10. The new drone, ...
- Robotic Research LLC to Exhibit at CES 2020on January 3, 2020 at 10:35 am
The company designs and builds unmanned transformable robots as well as robotic hardware/software and autonomy kits for autonomous vehicles and driver-supported systems that increase transportation ...
via Bing News