From bionic limbs to sentient robots, robotic entities in science fiction blur the boundaries between biology and machine. Realistic robots are far behind in comparison. While we will not get to the level Star Trek Data anytime soon, there is now a robot hand with a sense of touch that is almost human.
One thing that robots have not been able to achieve is a level of sensitivity and dexterity high enough to feel and manipulate objects as humans do. Enter a robotic hand developed by a team of researchers at Columbia University. (Five years ago, we covered their work back when this achievement was still understandable.)
This hand does not just pick things up and put them under orders. It is so sensitive that it can actually “feel” what it is touching, and is subtle enough to easily change the position of its fingers so it can better hold objects, a maneuver known as “toewalking”. He is so sensitive he can even do all this in the dark, as he detects everything by touch.
State space navigation
“[This is] The researchers said in a study recently published on the preprint server arXiv.
To create this hand, the Columbia team needed to find the most efficient way to navigate through the so-called state space structure. Each known possible configuration of a system is called its own state space. The state space structure describes how the robot is supposed to move from one step to the next within that state space. There are different machine learning methods that can train it to do this.
A popular method of training a robot is known as reinforcement learning (RL). This can be considered a “good bot” vs. a “bad bot” approach. The robot’s control software is “rewarded” for doing what it’s supposed to and “punished” for anything it does incorrectly. He learns through trial and error until he knows how he is supposed to act. Unfortunately, RL has its drawbacks because the slightest deviation from the expected state can cause the robot to drop an object.
So the team also used sampling-based planning (SBP) algorithms to give the robot a better grip (pun intended) on its space structure. The SPB does not need to go through every possible combination of motions to access the state space; Instead, it randomly samples different tracks. Each successful maneuver attempted by the robot is stored using the SBP as a new branch added to a digital tree, which the AI can refer to later when looking for a way to solve a problem. SBP still has its issues – it can only rely on what you’ve done before, and unexpected obstacles encountered in the state space can be problematic.
“[We used] The researchers said the strength of both RL and SBP methods for training motor control policies for hand manipulation with finger gait. “We aim to tackle the most difficult objects, including concave shapes, while securing them at all times without relying on supporting surfaces.”
Back to her senses
For AI, coming up with a set of directions is the easy part. You can tell a bot what to do, but most bots can’t provide much in the way of feedback. The new robot hand goes beyond that with fingers that can feel exactly what it is touching and sense the movement and location of an object. To do this, it needed another algorithm, the Rapid Reconnaissance Tree (RRT). This algorithm is behind the hand’s ability to handle even the most difficult things. RRT finds the branch of the tree that is the shortest path through the state space to the state that represents a completed task.
This combination of algorithms ended up making this robot hand like no other. The researchers taught it to keep at least three fingers in contact with the object and to balance the force used by each finger if an object began to slip or if its shape required different amounts of pressure to maintain its grip. Closed loop control has also been used to train the hand further by giving it feedback at various points throughout the process.
This robotic hand is just as adept in the dark as it is when it can “see” its surroundings, just as a human hand is when it’s trying to feel for something. This is allergic sensing, which many organisms are capable of. Because the hand can have such an amazing sense of touch, it could potentially be used as a more advanced form of assistance for people who need assistance with certain tasks.
We are still far away from androids like Data, who can sense anything. But we at least now have a robotic hand that’s sensitive and sensitive enough to literally stay in touch.
Elizabeth Raine Creature Writes. Her work has appeared on SYFY WIRE, Space.com, Live Science, Grunge, Den of Geek, and Forbidden Futures. When she’s not writing, she’s either shape-shifting, drawing, or masquerading as a character no one has ever heard of. Follow her on Twitter: @hravenrayne.
#Hypersensitive #Robot #Hand #Scary #Human #Feels #Ars #Technica