<p>This article originally appeared in <a href="https://news.stanford.edu/2023/10/03/robot-provides-personalized-room-cleanup/" target="_blank" rel="noopener noreferrer">Stanford University</a> and has been republished with permission.</p><hr><p><span contenteditable="false" draggable="true" class="fr-video fr-deletable fr-fvc fr-dvb fr-draggable"><iframe width="640" height="360" src="https://www.youtube.com/embed/dYdi0Aek664?&amp;wmode=opaque&amp;rel=0&amp;enablejsapi=1&amp;origin=https://www.wevolver.com" frameborder="0" allowfullscreen="" class="fr-draggable" data-gtm-yt-inspected-8="true" data-lf-form-tracking-inspected-kn9eq4roawkarlvp="true" data-lf-yt-playback-inspected-kn9eq4roawkarlvp="true" data-lf-vimeo-playback-inspected-kn9eq4roawkarlvp="true"><span class="fr-mk" style="display: none;">&nbsp;</span></iframe></span><em>The one-armed robot TidyBot can encounter new objects and determine – with about 85% accuracy – where they should be put away, based on the personalized preferences it’s been given previously.</em></p><p id="isPasted">Whenever <a href="https://profiles.stanford.edu/shuran-song" target="_blank">Shuran Song</a> teaches robotics, she asks her students the same question: What daily tasks would you most like a robot to do?</p><p>“The two requests that always come up are ‘washing dishes’ and ‘cleaning my room,’” said Song, assistant professor of electrical engineering at Stanford.</p><p>So far, the first task is best left to a dishwasher. But for the second, Song and an interinstitutional team of roboticists may have a solution: <a href="https://tidybot.cs.princeton.edu/" target="_blank">TidyBot</a>, a one-armed helper that cleans according to personal preferences. The group, which includes researchers from Princeton University, The Nueva School, Columbia University, and Google, presented a paper on the robot and its performance at the International Conference on Intelligent Robots and Systems on October 2nd. An extended journal version of <a href="https://arxiv.org/abs/2305.05658" target="_blank">the paper</a> will appear soon in <em>Autonomous Robots</em> in the special issue on robots and large language models.</p><p>“We all experience it: our home or lab or office gets messy,” said <a href="https://profiles.stanford.edu/jeannette-bohg" target="_blank">Jeannette Bohg</a>, assistant professor of computer science at Stanford and another of the paper’s authors. “Now imagine you just tell a robot, ‘In my place, this is where that thing goes, and that is where this thing goes.’ Then you leave and let the robot tidy up.”</p><h2>For personal pick-up, no one size fits all</h2><p>To navigate your space, TidyBot rolls around on powered wheels hidden beneath a box-shaped platform. A seven-jointed arm sprouting from the top ends in a two-fingered gripper for manipulating objects, from opening a drawer and gently placing a toy inside to precisely tossing an empty can into a bin. Cameras on the ceiling allow TidyBot to pinpoint objects and obstacles across a room, while others on its body provide close-up views to analyze items in front of it.</p><p>But it’s TidyBot’s adaptability, not just its physical form, that could make it so useful around the house.</p><p>“Household organization is so personal,” Bohg said. Recognizing this, the group worked to make TidyBot customizable to different preferences.</p><p>TidyBot uses a large language model (a type of AI similar to the one used in ChatGPT) trained on massive amounts of information from the internet to identify broad categories of objects like clothes, food, and toys. This means you can give TidyBot just a few examples of where certain belongings go and it will understand to clean up similar items in the same ways – even if it’s never seen those items before.</p><p>Suppose, for instance, that you want your kid’s loose toys back on their shelf. “You could say, ‘I want the teddy bear on the shelf,’ and TidyBot is able to generalize that concept to other toys in your home,” Song explained. TidyBot also recognizes qualities like color, allowing it to do things like sort light and dark laundry. In real-world tests, the robot can properly put away 85% of objects.</p><p><span class="fr-img-caption fr-fic fr-dib" style="width: 768px;"><span class="fr-img-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjk2NTA2ODYwNzEyLXRpZHlib3Rfc3JfMi53ZWJwIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" style="width: 766px;" class="fr-fic fr-dib"><span class="fr-inner">TidyBot is trained within a specific space so it can learn how different types of objects are put away in that area. For example, the researchers were able to make it put clothes on the couch. (Image credit: Kurt Hickman)</span></span></span></p><h2 id="isPasted">An endearingly imperfect helper</h2><p>That 85% success rate means TidyBot still makes mistakes, of course – sometimes to comical effect. Bohg recalls a visit from a former adviser during a TidyBot demo.</p><p>“He was visible in TidyBot’s ‘playing field,’ and TidyBot tried to pick him up,” she laughed.</p><p>Luckily, it didn’t succeed: the team designed the robot mainly for handheld-sized objects or soft, deformable items like clothing. TidyBot also struggles with difficult-to-grasp shapes, like a credit card lying flat on a table. Bohg hopes to improve robots’ abilities to replicate human strategies in scenarios like this – such as dragging the card to the table’s edge to pick it up. And she’d like to investigate whether multiple TidyBots could cooperate to accomplish bigger tasks like moving furniture or making a bed.</p><p>Like most humans, TidyBot also has more trouble with new spaces than with new objects: an unfamiliar kind of drawer in a strange location might trip it up, just as it might take time to learn your way around a new kitchen.</p><p>“The challenge is having TidyBot work in all sorts of environments – in your home and my home, in this lab and the other lab,” Bohg said.</p><p>Getting TidyBot to generalize across environments as well as objects will take time and effort. But as large language models and vision models improve, so will TidyBot’s ability to process information and make correct decisions in new scenarios. In theory, Song said, there may be no limit to how many combinations of items and environments TidyBot could eventually work with. As long as an object exists on the internet, TidyBot could learn to recognize and categorize it.</p><p>“Maybe in the long term we can think about a more general-purpose robot that’s also able to cook my meal, make my bed, do my laundry,” Song mused. But for now, she said, “If I have a robot that’s able to reliably put my things away, that’s a pretty good goal to achieve already.”</p>

TidyBot is trained within a specific space so it can learn how different types of objects are put away in that area. For example, the researchers were able to make it put clothes on the couch. (Image credit: Kurt Hickman)

The one-armed “TidyBot” needs only a few starter examples to figure out how you like your stuff put away.

This robot learns to clean your space just the way you like it

Part 2 of our dexterous robotic hand series. 

Dexterous Robotic Hands Part 2: How the Shadow Dexterous Hand is Revolutionizing Robotics

<p id="isPasted">There are many robot arms on the market, and you&rsquo;re probably familiar with a few. Your modern cobot ranges in cost from &euro;25k to &euro;200k &ndash; and that is for something with 6 or 7 degrees of freedom (DoF). So you might say &ldquo;&euro;4k-&euro;14k per DoF&rdquo; and use that as a &ldquo;rule of thumb&rdquo; to price off-the-shelf commercial robots.</p><p>There are plenty of people with very low-cost robot hands and arms out there &ndash; if you&rsquo;re a hobbyist or you love to prototype, then that&rsquo;s where you should go! A quick look online finds 3d printed structures for a few tens or hundreds of Euro to print, although you still have to supply valves and compressors and electronics and maybe sensing. A fun project for a hardware developer! Then there are some cool hobbyist complete systems &ndash; I can find a 6DoF robot arm on Amazon for &euro;120 today, which would be great for teaching beginners the fundamentals of robot operation.</p><p>But if you&rsquo;re planning to do work with the robot, then you want a complete system. And you want something that will lift a few kilos, has the sensors and motors to move with precision and speed across a decent set of kinematics, has safety circuits and controls built in, and supports the &ldquo;mechanical engineering&rdquo; interfaces allowing you to build it into your lab properly. Suddenly you&rsquo;re back to &ldquo;it&rsquo;s going to be a 5 figure sum&rdquo;. And there are good reasons for that &ndash; you&rsquo;re getting something with mechanical strength, highly engineered sensors, motors and gearboxes, and sophisticated control systems to make it work.</p><p><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjk0NzcwMTg4NzE4LVVudGl0bGVkICgyOTcgw5cgMjEwbW0pICgyMDAgw5cgMjAwbW0pICgzMDAgw5cgMTAwbW0pLnBuZyIsImVkaXRzIjp7InJlc2l6ZSI6eyJ3aWR0aCI6OTUwLCJmaXQiOiJjb3ZlciJ9fX0=" style="width: 766px;" class="fr-fic fr-dib"><br></p><p>And then you&rsquo;re going to want documentation, software interfaces, and support when it doesn&rsquo;t do what you want.</p><p>Don&rsquo;t get me wrong &ndash; that isn&rsquo;t because you bought something that wasn&rsquo;t up to the job. Most support tickets in robotics come with one of two sentences at the start: &ldquo;I was doing &hellip; and it broke&rdquo; or &ldquo;How can I get it to do &hellip;&rdquo;. And we&rsquo;ve seen enough robot arms slam things into tables to know that breaking things is part of the game here &ndash; honestly, if you&rsquo;re not breaking the odd robot you&rsquo;re probably not doing research at the frontiers!</p><p>So that&rsquo;s a real, commercial robot arm. Assume &euro;5k per DoF for a generic cobot, and maybe &euro;8-10k if you&rsquo;re getting something really amazing. Even the dinky 6DoF arm we use in one of our production processes was &euro;17k for 6DoF (to be fair to it, the precision is really impressive, but it&rsquo;s still shoebox-size).</p><p>However, <a href="https://www.shadowrobot.com/" target="_blank" rel="noopener noreferrer">Shadow Robot</a> isn&#39;t in the robot arm business. We make what the marketing people politely call &ldquo;end of arm tooling&rdquo;. This space ranges from welding guns to lasers, from two jaw grippers to vacuum suckers, and right the way over there &mdash;&gt; robot hands. And according to the good people at the International Federation of Robotics, adding EoAT and the other parts needed to make it work can triple the price of a robot arm!</p><p>A simple gripper that opens and closes without any fine movement control, without force sensing and without position sensing might well be included in the price of the cobot. Lots of companies do that and it works well for their customers. But we&rsquo;re here for data. We want precise control. We want to be able to move each fingertip independently and position it accurately in space with respect to the other fingertips.</p><p>In short, we want a robot hand where each finger behaves like a small robot arm &ndash; and we want all that complexity and performance in something 10x smaller! What would that cost? Well, &euro;5k per DoF gets us to &euro;100k for 20DoF &hellip; and the price for the full Shadow Hand Plus today is &hellip; &euro;110k including shipping, installation, training and support!</p><p>Of course, we can make simpler Hands with fewer degrees of freedom &ndash; and fewer fingers &nbsp;&ndash; and these go right down to a 2-finger 7DoF system which is more like &euro;45k including shipping, installation, training and support. You can see the full range <a href="https://www.shadowrobot.com/dexterous-hand-series/" target="_blank">here</a>, and if you want to talk to us about getting one, then <a href="https://www.shadowrobot.com/contact/" target="_blank" rel="noopener noreferrer">get in touch</a>!&nbsp;</p>

Spending your hard-earned budget on a robot is a step you don’t want to take lightly, Any worthwhile robot is likely to represent a significant piece of capital equipment.

So, let’s take a look at some hardware and what you might expect to pay for.

How Much does a Robot Hand Cost?

Agile, robust, and dexterous hands are a challenging puzzle piece to the potential of robotics. 


Dexterous Robotic Hands Part 1: Unraveling the History, Technology, and Applications of Dexterous Robotic Hands

In this third part of the series, we will take a deeper look at the development of the Shadow Hand and Glove, the technology behind the glove, and the potential impact of this system on the field of robotics.


Dexterous Robotic Hands Part 3: Shadow Gloves - The Final Piece of the Robotics Puzzle

<p id="isPasted">With their ability to process vast amounts of text data, and to use this information to answer prompts, neural networks known as large language models (LLMs) like Chat-GPT have been making headlines for their potential to change the way we <a href="https://longread.epfl.ch/en/dossier/artificial-intelligence-friend-or-foe/" target="_blank">write, learn, and even make art</a>. Now, EPFL researchers have applied the technology to a new sphere: robotic design.</p><p>In <a href="https://www.nature.com/articles/s42256-023-00669-7" rel="noopener" target="_blank">a case study</a> published in <em>Nature Machine Intelligence</em>, Josie Hughes, head of the <a href="https://www.epfl.ch/labs/create/" target="_blank">Computational Robot Design &amp; Fabrication Lab</a> in the School of Engineering, EPFL PhD student Francesco Stella, and Cosimo Della Santina of TU Delft used Chat-GPT to design a working robotic tomato-harvester. The study provides a framework for humans and LLMs to design such devices collaboratively. Based on their experience, the researchers describe opportunities and risks of applying artificial intelligence (AI) tools to robotics, which they argue &ldquo;could change the way we design robots, while enriching and simplifying the process.&rdquo;</p><p><span class="fr-img-caption fr-fic fr-dib" style="width: 768px;"><span class="fr-img-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjg3MTc5NTUyMzcwLTM0Njh4NDYyNC5qcGciLCJlZGl0cyI6eyJyZXNpemUiOnsid2lkdGgiOjk1MCwiZml0IjoiY292ZXIifX19" style="width: 766px;" class="fr-fic fr-dib"><span class="fr-inner">&copy; CREATE Lab/EPFL&nbsp;</span></span></span></p><p>&ldquo;Even though Chat-GPT is a language model and its code generation is text-based, it provided significant insights and intuition for physical design, and showed great potential as a sounding board to stimulate human creativity,&rdquo; says Hughes.</p><p><strong>Potential and pitfalls of AI as &lsquo;inventor&rsquo;</strong></p><p>In a first phase, the researchers and LLM engaged in an &lsquo;ideation&rsquo; discussion to define their robot&rsquo;s purpose, design parameters, and specifications. A second phase was devoted to realizing the robot in the real world, which involved refining the LLM-generated code, fabricating the device, and troubleshooting its functioning.</p><p><span class="fr-img-caption fr-fic fr-dib" style="width: 768px;"><span class="fr-img-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjg3MTc5NTc2MzA5LTMwMjR4NDAzMi5qcGciLCJlZGl0cyI6eyJyZXNpemUiOnsid2lkdGgiOjk1MCwiZml0IjoiY292ZXIifX19" style="width: 766px;" class="fr-fic fr-dib"><span class="fr-inner">&copy; CREATE Lab/EPFL&nbsp;</span></span></span></p><p>For the first phase, the researchers started at a high conceptual level, conversing with the LLM on future challenges to humanity, and identifying robotic crop harvesting as a solution to the challenge of global food supply. They then drew upon the LLM&rsquo;s access to global data from academic publications, technical manuals, books, and media to provide the &ldquo;most probable&rdquo; answer to prompts such as &ldquo;what features should a robot harvester have?&rdquo;</p><p>Once a basic robotic format was identified (a motor-driven gripper for grasping ripe tomatoes), the researchers could then pose more specific questions, like &ldquo;what shape should the gripper have?&rdquo;, and ask the LLM to make technical suggestions including materials and computer code for controlling the device.</p><p>&ldquo;While computation has been largely used to assist engineers with technical implementation, for the first time, an AI system can ideate new systems, thus automating high-level cognitive tasks. This could involve a shift of human roles to more technical ones,&rdquo; Stella says.</p><p><span class="fr-img-caption fr-fic fr-dib" style="width: 768px;"><span class="fr-img-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjg3MTc5NzE2MjgzLWV6Z2lmLTUtNTdiZWU1MWE4Zi5qcGciLCJlZGl0cyI6eyJyZXNpemUiOnsid2lkdGgiOjk1MCwiZml0IjoiY292ZXIifX19" style="width: 766px;" class="fr-fic fr-dib"><span class="fr-inner">&copy; CREATE Lab/EPFL&nbsp;</span></span></span></p><p>In addition to assigning Chat-GPT the role of &lsquo;inventor&rsquo;, the researchers outlined other possible human-LLM collaboration modes in their paper. For example, &lsquo;collaborative exploration&rsquo; uses AI to augment researchers&rsquo; expertise by contributing wide-ranging knowledge beyond their own fields. AI can also act as a &lsquo;funnel&rsquo;, helping to refine the design process and providing technical input, with humans retaining creative control.</p><p>As there are logical and ethical risks associated with each collaboration mode, the researchers caution that the role of LLMs must be carefully evaluated going forward. For example, the use of LLMs raises questions of bias, plagiarism, and intellectual property, as it&rsquo;s unclear if an LLM-generated design can be considered novel.</p><p>&ldquo;In our study, Chat-GPT identified tomatoes as the crop &lsquo;most worth&rsquo; pursuing for a robotic harvester. However, this may be biased towards crops that are more covered in literature, as opposed to those where there is truly a real need. When decisions are made outside the scope of knowledge of the engineer, this can lead to significant ethical, engineering, or factual errors.&rdquo; Hughes says.</p><p><span class="fr-img-caption fr-fic fr-dib" style="width: 768px;"><span class="fr-img-wrap"><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1687180178832-ezgif-5-62f982faad.jpg" style="width: 766px;" class="fr-fic fr-dib"><span class="fr-inner">&copy; CREATE Lab/EPFL&nbsp;</span></span></span></p><p>Despite these cautions, Hughes and her team conclude, based on their experience, that LLMs have great potential to be a force for good, if well managed: &ldquo;The robotics community must therefore identify how to leverage these powerful tools to accelerate the advancement of robots in an ethical, sustainable, and socially empowering way.&rdquo;</p>

EPFL researchers have used Chat-GPT-3 to develop a robotic gripper for harvesting tomatoes, in a first demonstration of the artificial intelligence tool’s potential for collaborating with humans on robot design.

Researchers unveil first Chat-GPT-designed robot

<p id="isPasted">Grasping objects of different sizes, shapes and textures is a problem that is easy for a human, but challenging for a robot. Researchers from the University of Cambridge designed a soft, 3D printed robotic hand that cannot independently move its fingers but can still carry out a range of complex movements.&nbsp;</p><p>The robot hand was trained to grasp different objects and was able to predict whether it would drop them by using the information provided from sensors placed on its &lsquo;skin&rsquo;.&nbsp;</p><p>This type of passive movement makes the robot far easier to control and far more energy-efficient than robots with fully motorised fingers. The researchers say their adaptable design could be used in the development of low-cost robots that are capable of more natural movement and can learn to grasp a wide range of objects. The <a href="https://onlinelibrary.wiley.com/doi/full/10.1002/aisy.202200390" target="_blank">results</a> are reported in the journal <em>Advanced Intelligent Systems</em>.&nbsp;</p><p><span class="fr-video fr-deletable fr-fvc fr-dvb fr-draggable" contenteditable="false" draggable="true"><iframe width="640" height="360" src="https://www.youtube.com/embed/haWziO6aOqc?&wmode=opaque&rel=0&enablejsapi=1&origin=https://www.wevolver.com" frameborder="0" allowfullscreen="" class="fr-draggable" data-lf-form-tracking-inspected-kn9eq4roawkarlvp="true" data-lf-yt-playback-inspected-kn9eq4roawkarlvp="true" data-lf-vimeo-playback-inspected-kn9eq4roawkarlvp="true"></iframe></span></p><p id="isPasted">In the natural world, movement results from the interplay between the brain and the body: this enables people and animals to move in complex ways without expending unnecessary amounts of energy. Over the past several years, soft components have begun to be integrated into robotics design thanks to advances in 3D printing techniques, which have allowed researchers to add complexity to simple, energy-efficient systems.&nbsp;</p><p>The human hand is highly complex, and recreating all of its dexterity and adaptability in a robot is a massive research challenge. Most of today&rsquo;s advanced robots are not capable of manipulation tasks that small children can perform with ease. For example, humans instinctively know how much force to use when picking up an egg, but for a robot this is a challenge: too much force, and the egg could shatter; too little, and the robot could drop it. In addition, a fully actuated robot hand, with motors for each joint in each finger, requires a significant amount of energy.&nbsp;</p><p>In <a href="http://www.eng.cam.ac.uk/profiles/fi224" target="_blank">Professor Fumiya&nbsp;</a><a href="http://www.eng.cam.ac.uk/profiles/fi224" target="_blank">Iida</a>&rsquo;s <a href="https://birlab.org/" target="_blank">Bio-Inspired Robotics Laboratory</a> in Cambridge&rsquo;s Department of Engineering, researchers have been developing potential solutions to both problems: a robot hand than can grasp a variety of objects with the correct amount of pressure while using a minimal amount of energy.&nbsp;</p><p>&ldquo;In <a href="http://www.eng.cam.ac.uk/news/3d-printed-robot-hand-plays-piano" target="_blank">earlier experiments</a>, our lab has shown that it&rsquo;s possible to get a significant range of motion in a robot hand just by moving the wrist,&rdquo; said co-author Dr Thomas George-Thuruthel, who is now based at University College London (UCL) <a href="https://www.ucl.ac.uk/ucl-east/study-and-research/robotics-and-autonomous-systems" target="_blank">East</a>. &ldquo;We wanted to see whether a robot hand based on passive movement could not only grasp objects, but would be able to predict whether it was going to drop the objects or not, and adapt accordingly.&rdquo;&nbsp;</p><p>The researchers used a 3D-printed anthropomorphic hand implanted with tactile sensors, so that the hand could sense what it was touching. The hand was only capable of passive, wrist-based movement.&nbsp;</p><p>The team carried out more than 1,200 tests with the robot hand, observing its ability to grasp small objects without dropping them. The robot was initially trained using small 3D printed plastic balls, and grasped them using a pre-defined action obtained through human demonstrations.&nbsp;</p><p>&ldquo;This kind of hand has a bit of springiness to it: it can pick things up by itself without any actuation of the fingers,&rdquo; said first author Dr Kieran Gilday, who is now based at EPFL in Lausanne, Switzerland. &ldquo;The tactile sensors give the robot a sense of how well the grip is going, so it knows when it&rsquo;s starting to slip. This helps it to predict when things will fail.&rdquo;&nbsp;</p><p>The robot used trial and error to learn what kind of grip would be successful. After finishing the training with the balls, it then attempted to grasp different objects including a peach, a computer mouse and a roll of bubble wrap. In these tests, the hand was able to successfully grasp 11 of 14 objects.&nbsp;</p><p><span class="fr-img-caption fr-fic fr-dib" style="width: 768px;"><span class="fr-img-wrap"><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1684333513339-CENTRE_RobotHand_CREDIT+Bio-Inspired+Robotics+Laboratory%2C+University+of+Cambridge.jpg" style="width: 766px;" class="fr-fic fr-dib"><span class="fr-inner">Credit: Bio-Inspired Robotics Laboratory, University of Cambridge.&nbsp;</span></span></span></p><p id="isPasted">&ldquo;The sensors, which are sort of like the robot&rsquo;s skin, measure the pressure being applied to the object,&rdquo; said Dr George-Thuruthel. &ldquo;We can&rsquo;t say exactly what information the robot is getting, but it can theoretically estimate where the object has been grasped and with how much force.&rdquo;&nbsp;</p><p>&ldquo;The robot learns that a combination of a particular motion and a particular set of sensor data will lead to failure, which makes it a customisable solution,&rdquo; said Dr Gilday. &ldquo;The hand is very simple, but it can pick up a lot of objects with the same strategy.&rdquo;&nbsp;</p><p>&ldquo;The big advantage of this design is the range of motion we can get without using any actuators,&rdquo; said Professor Iida. &ldquo;We want to simplify the hand as much as possible. We can get lots of good information and a high degree of control without any actuators, so that when we do add them, we&rsquo;ll get more complex behaviour in a more efficient package.&rdquo;&nbsp;</p><p>A fully actuated robotic hand, in addition to the amount of energy it requires, is also a complex control problem. The passive design of the Cambridge-designed hand, using a small number of sensors, is easier to control, provides a wide range of motion, and streamlines the learning process.&nbsp;</p><p>In future, the system could be expanded in several ways, such as by adding computer vision capabilities, or teaching the robot to exploit its environment, which would enable it to grasp a wider range of objects.&nbsp;</p><hr><p><strong id="isPasted">Reference:</strong><br><em>K. Gilday, Dr. T. George-Thuruthel, Prof. F. Iida. &lsquo;<a href="https://onlinelibrary.wiley.com/doi/full/10.1002/aisy.202200390" target="_blank">Predictive Learning of Error Recovery with a Sensorised Passivity-based Soft Anthropomorphic Hand</a>.&rsquo; Advanced Intelligent Systems (2023). 10.1002/aisy.202200390</em>&nbsp;</p>

Researchers have designed a low-cost, energy-efficient robotic hand that can grasp a range of objects – and not drop them – using just the movement of its wrist and the feeling in its ‘skin’.  

It's all in the wrist: energy-efficient robot hand learns how not to drop the ball

<section id="isPasted"><p>Think about what you do with your hands when you&rsquo;re home at night pushing buttons on your TV&rsquo;s remote control, or at a restaurant using all kinds of cutlery and glassware. These skills are all based on touch, while you&rsquo;re watching a TV program or choosing something from the menu. Our hands and fingers are incredibly skilled mechanisms, and highly sensitive to boot.</p><p>Robotics researchers have long been trying to create &ldquo;true&rdquo; dexterity in robot hands, but the goal has been frustratingly elusive. Robot grippers and suction cups can pick and place items, but more dexterous tasks such as assembly, insertion, reorientation, packaging, etc. have remained in the realm of human manipulation. However, spurred by advances in both sensing technology and machine-learning techniques to process the sensed data, the field of robotic manipulation is changing very rapidly.</p></section><section><h2>Highly dexterous robot hand even works in the dark</h2><p><span class="fr-video fr-deletable fr-fvc fr-dvb fr-draggable" contenteditable="false" draggable="true"><iframe width="640" height="360" src="https://www.youtube.com/embed/mYlc_OWgkyI?&wmode=opaque&rel=0&enablejsapi=1&origin=https://www.wevolver.com" frameborder="0" allowfullscreen="" class="fr-draggable" data-lf-form-tracking-inspected-kn9eq4roawkarlvp="true" data-lf-yt-playback-inspected-kn9eq4roawkarlvp="true" data-lf-vimeo-playback-inspected-kn9eq4roawkarlvp="true"></iframe></span><em>Video Credit: Columbia University ROAM Lab&nbsp;</em></p><section id="isPasted"><p>Researchers at<a href="http://engineering.columbia.edu/" rel="nofollow" target="_blank">&nbsp;Columbia Engineering</a> have demonstrated a highly dexterous robot hand, one that combines an advanced sense of touch with motor learning algorithms in order to achieve a high level of dexterity.&nbsp;</p><p>As a demonstration of skill, the team chose a difficult manipulation task: executing an arbitrarily large rotation of an unevenly shaped grasped object in hand while always maintaining the object in a stable, secure hold. This is a very difficult task because it requires constant repositioning of a subset of fingers, while the other fingers have to keep the object stable. Not only was the hand able to perform this task, but it also did it without any visual feedback whatsoever, based solely on touch sensing.&nbsp;</p><p>In addition to the new levels of dexterity, the hand worked without any external cameras, so it&#39;s immune to lighting, occlusion, or similar issues.&nbsp;And the fact that the hand does not rely on vision to manipulate objects means that it can do so in very difficult lighting conditions that would confuse vision-based algorithms--it can even operate in the dark.</p><p>&ldquo;While our demonstration was on a proof-of-concept task, meant to illustrate the capabilities of the hand, we believe that this level of dexterity will open up entirely new applications for robotic manipulation in the real world,&rdquo; said <a href="https://engineering.columbia.edu/faculty/matei-ciocarlie" rel="nofollow" target="_blank">Matei Ciocarlie</a>, associate professor in the Departments of<a href="https://me.columbia.edu/" rel="nofollow" target="_blank">&nbsp;Mechanical Engineering</a> and<a href="http://www.cs.columbia.edu/" rel="nofollow" target="_blank">&nbsp;Computer Science</a>. &ldquo;Some of the more immediate uses might be in logistics and material handling, helping ease up supply chain problems like the ones that have plagued our economy in recent years, and in advanced manufacturing and assembly in factories.&rdquo;</p></section><section><h2>Leveraging optics-based tactile fingers</h2><p><span class="fr-img-caption fr-fic fr-dib" style="width: 768px;"><span class="fr-img-wrap"><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1682949784160-optics-based-robot-fingers.jpg" style="width: 766px;" class="fr-fic fr-dib"><span class="fr-inner">Credit: Columbia University ROAM Lab&nbsp;</span></span></span></p><section id="isPasted"><p>In earlier work, Ciocarlie&rsquo;s group collaborated with Ioannis Kymissis, professor of electrical engineering, to develop a new generation of <a href="https://www.engineering.columbia.edu/news/ciocarlie-tactile-robot-finger" rel="nofollow" target="_blank">optics-based tactile robot fingers</a>. These were the first robot fingers to achieve contact localization with sub-millimeter precision while providing complete coverage of a complex multi-curved surface. In addition, the compact packaging and low wire count of the fingers allowed for easy integration into complete robot hands.&nbsp;</p></section><section><h2>Teaching the hand to perform complex tasks</h2><p><span class="fr-img-caption fr-fic fr-dib" style="width: 768px;"><span class="fr-img-wrap"><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1682949808021-robot-hand-holding-unevenly-shaped-object.jpg" style="width: 766px;" class="fr-fic fr-dib"><span class="fr-inner">Credit: Columbia University ROAM Lab&nbsp;</span></span></span></p><section id="isPasted"><p>For this new work, led by CIocarlie&rsquo;s doctoral researcher, Gagan Khandate, the researchers designed and built a robot hand with five fingers and 15 independently actuated joints--each finger was equipped with the team&rsquo;s touch-sensing technology. The next step was to test the ability of the tactile hand to perform complex manipulation tasks. To do this, they used new methods for motor learning, or the ability of a robot to learn new physical tasks via practice. In particular, they used a method called deep reinforcement learning, augmented with new algorithms that they developed for effective exploration of possible motor strategies.&nbsp;</p></section><section><h2>Robot completed approximately one year of practice in only hours of real-time</h2><p>The input to the motor learning algorithms consisted exclusively of the team&rsquo;s tactile and proprioceptive data, without any vision. Using simulation as a training ground, the robot completed approximately one year of practice in only hours of real-time, thanks to modern physics simulators and highly parallel processors. The researchers then transferred this manipulation skill trained in simulation to the real robot hand, which was able to achieve the level of dexterity the team was hoping for. Ciocarlie noted that &ldquo;the directional goal for the field remains assistive robotics in the home, the ultimate proving ground for real dexterity. In this study, we&#39;ve shown that robot hands can also be highly dexterous based on touch sensing alone. Once we also add visual feedback into the mix along with touch, we hope to be able to achieve even more dexterity, and one day start approaching the replication of the human hand.&rdquo;</p></section><section><h2>Ultimate goal: joining abstract intelligence with embodied intelligence</h2><p>Ultimately, Ciocarlie observed, a physical robot being useful in the real world needs both abstract, semantic intelligence (to understand conceptually how the world works), and embodied intelligence (the skill to physically interact with the world). Large language models such as OpenAI&rsquo;s GPT-4 or Google&rsquo;s PALM aim to provide the former, while dexterity in manipulation as achieved in this study represents complementary advances in the latter. &nbsp;</p><p>For instance, when asked how to make a sandwich, ChatGPT will type out a step-by-step plan in response, but it takes a dexterous robot to take that plan and actually make the sandwich. In the same way, researchers hope that physically skilled robots will be able to take semantic intelligence out of the purely virtual world of the Internet, and put it to good use on real-world physical tasks, perhaps even in our homes.</p><p>The paper has been accepted for publication at the&nbsp;<a href="https://roboticsconference.org/" rel="nofollow" target="_blank">Robotics: Science and Systems Conference</a> (Daegu, Korea, July 10-14, 2023), and is currently <a href="https://urldefense.proofpoint.com/v2/url?u=https-3A__arxiv.org_abs_2303.03486&d=DwMFaQ&c=009klHSCxuh5AI1vNQzSO0KGjl4nbi2Q0M1QLJX9BeE&r=N_SVJ4CjfYnqNc0c8S28HoUw1f1W7VxuRSEGf5V1elA&m=tnipN5N6mn9VlePBq2FZxfWCw8P7d1wGFxiy5uKMCsIsw8hROX6SfydHukNDOLSP&s=Rla29TNWbWzsHyimr-lmaCHpcAHLQttUzI8CMAOsCY8&e=" rel="nofollow" target="_blank">available as a preprint</a>.</p></section></section></section></section>

Columbia Engineers design a robot hand that is the first device of its kind to join advanced sense of touch with motor-learning algorithms — it doesn’t rely on vision to manipulate objects

Highly Dexterous Robot Hand Can Operate in the Dark - Just Like Us

Clone Robotics’ humanoid hand uses hydraulics to mimic complex human hand function. 


Could this lifelike robotic hand be the start of a humanoid future?

<section id="isPasted"><p><strong>A cleverly designed pressure valve lets soft robots respond to their environment without computer control. How that works, is what researchers from Eindhoven University of Technology (TU/e) explain in a recently published article in the journal Matter. The research by <a href="https://www.tue.nl/en/research/researchers/bas-overvelde/" target="_blank">Bas Overvelde</a>, associate professor at the department of&nbsp;<a href="https://www.tue.nl/en/our-university/departments/mechanical-engineering/research/" target="_blank">Mechanical Engineering</a>, and PhD researcher Luuk van Laake, brings robots that move and feel like living beings closer. Such designs are better suited for exploring rough and unfamiliar terrain or for medical applications.</strong></p></section><section tabindex="-1"><p>Robots are still mainly associated with hard machines, controlled by a central computer that has to think about each step. Living beings, on the other hand, move smoothly because intelligent behavior is embedded in their bodies, which is also ideal for robots that constantly interact with people, such as in medical care. The research field of&nbsp;<a href="https://www.tue.nl/en/news-and-events/news-overview/01-06-2022-collaborating-on-a-real-life-barbapapa/" target="_blank">soft robotics</a> therefore works on robots made of soft, flexible materials that respond to changes in their environment without external control.</p><p>Bas Overvelde, who besides his work at TU/e is also part of&nbsp;<a href="https://overvelde.com/" target="_blank">AMOLF</a>: &quot;We want to make robots without a central computer, that move and react to their environment thanks to built-in reflexes in the robot body.&quot; In the journal Matter, Overvelde&#39;s team presents such a soft robot that operates on air pressure (without electronics), which walks and changes rhythm by responding to the environment thanks to a cleverly designed valve.</p><h2><strong>Ketchup bottle</strong></h2><p>The heart of the new soft robot is a &#39;hysteretic valve&#39;, as the researchers call their invention in their professional publication. Outsiders may recognize the valve as a variation on the opening of a ketchup bottle. &quot;That ensures that you can easily dose ketchup and that the liquid doesn&#39;t leak&quot;, says Overvelde: &quot;Yet the ketchup sometimes sputters if you hold it upside down and squeeze it hard.&quot; The opening then quickly clatters open and closed. Modern bottles have been modified so that they sputter as little as possible, but Overvelde and colleagues wondered whether this normally undesirable behavior could be used to make a robot move.</p><p><span class="fr-video fr-deletable fr-fvc fr-dvb fr-draggable" contenteditable="false" draggable="true"><iframe src="https://www.youtube.com/embed/wHYL4cFTAY0?&wmode=opaque&rel=0&enablejsapi=1&origin=https://www.wevolver.com" frameborder="0" allowfullscreen="" class="fr-draggable" data-lf-yt-playback-inspected-kn9eq4roawkarlvp="true" style="width: 767px; height: 436px;"></iframe></span><em>Watch how the &#39;hysteretic valve&#39; works in this video.</em></p><p id="isPasted">In a computer model, the researchers first designed properties such as the stiffness of the tiny flaps in the valve opening. Then the valves were made by pouring silicone rubber into an accurate 3D printed mold. Using a laser, they cut tiny notches in it. This creates a valve that is normally closed, suddenly opens when the pressure is high enough, and only closes again when the pressure has dropped a bit. If you connect this valve to a pump and reservoir, a rhythm of pressure building up and air flowing through, automatically occurs. In this way the muscles of the soft robot are alternately inflated and relaxed.</p><p>The new component seems simple, but turned out to have hidden powers, says PhD researcher Luuk van Laake. &quot;When we made a computer model of two such valves connected to the same reservoir, they appeared to move in alternating directions very precisely. Completely unexpected, but it then turned out to actually work.&quot; This allowed the researchers to build a four-legged robot with a natural walking rhythm, without external commands. A soft grasping hand with five rhythmically drumming fingers also proved feasible.</p><h2><strong>Responding to the environment</strong></h2><p>The ultimate dream of soft robot designers is a robot that not only move by itself, but also reacts to its environment. Van Laake: &quot;Ultimately you want a robot that automatically changes its walking rhythm when it encounters an obstacle, purely because the muscles and joints respond to the counter-pressure.&quot;</p><p>That response to the environment is already quite successful in the lab. If the researchers briefly close an air hose by squeezing it, the pressure in the robot body changes. The robot then switches to a different walking rhythm. &quot;Our soft robot can respond to the environment without needing commands from a central computer&quot;, Van Laake explains.</p><h2><strong>Efficient</strong></h2><p>Before the soft robot is at the care bed, there is still a way to go. But the prospects are good, the researchers think. Overvelde: &quot;You can see here how you can make complex behavior with simple parts, without a computer that has every movement pre-made.&quot;</p><p>A soft robot that moves by built-in natural reflexes not only saves computing power, but also energy, according to Van Laake. &quot;In living creatures, you also see that a lot of movement and behavior comes from the construction of the body and the mechanics of muscles, rather than being prescribed from the brain.&quot; For example, our heart automatically balances the pressure in the left and right ventricles and we reuse the elastic energy stored in our tendons with each step as we walk. &quot;That turns out to be tremendously efficient. There&#39;s still a lot of behavior to explore in that regard.&quot;</p><h2><strong>More info</strong></h2><p>Lucas van Laake <em>et</em>. al. , <em><a href="https://www.sciencedirect.com/science/article/pii/S2590238522002971" target="_blank">A fluidic relaxation oscillator for reprogrammable sequential actuation in soft robots</a></em>, <em>Matter</em>, (2022).</p><hr><p>This article was first published here:&nbsp;<a href="https://www.tue.nl/en/news-and-events/news-overview/01-01-1970-sputtering-ketchup-bottle-lets-soft-robots-run-smoothly/" id="isPasted" rel="noopener noreferrer" target="_blank">https://www.tue.nl/en/news-and-events/news-overview/01-01-1970-sputtering-ketchup-bottle-lets-soft-robots-run-smoothly/</a></p></section>

The heart of the new soft robot is a 'hysteretic valve', as the researchers call their invention in a publication in the journal Matter.

Sputtering ketchup bottle lets soft robots run smoothly

<p>Mark R. Cutkosky is the Fletcher Jones Professor in the Dept. of Mechanical Engineering at Stanford University. He joined Stanford in 1985, after working in the Robotics Institute at Carnegie Mellon University and as a design engineer at ALCOA, in Pittsburgh, PA. He received his Ph.D. in Mechanical Engineering from Carnegie Mellon University in 1985.</p><p>Cutkosky’s research activities include robotic manipulation and tactile sensing and the design and fabrication of biologically inspired robots. He has graduated 48 Ph.D. students and published extensively in these areas. He consults with companies on robotics and human/computer interaction devices and holds several patents on related technologies. His work has been featured in Discover Magazine, The New York Times, National Geographic, Time Magazine and other publications and has appeared on PBS NOVA, CBS Evening News, and other popular media.</p><p>Cutkosky’s awards include a Fulbright Faculty Chair (Italy 2002), Fletcher Jones and Charles M. Pigott Chairs at Stanford University, an NSF Presidential Young Investigator award and Times Magazine Best Innovations (2006) for the Stickybot gecko-inspired robot. He is a fellow of ASME and IEEE and a member of Sigma Xi.</p><h1 dir="ltr">Soft Robotics With Mark Cutkosky&nbsp;</h1><p dir="ltr">Marwa ElDiwiny, host of the Soft Robotics podcast, interviews Mark about his work in soft robotics.</p><p dir="ltr"><span class="fr-video fr-deletable fr-fvc fr-dvb fr-draggable" contenteditable="false" draggable="true"><iframe src="https://open.spotify.com/embed-podcast/episode/6237w6uZUttAH7nW4lhAPf" width="100%" height="232" frameborder="0" allowtransparency="true" class="fr-draggable"></iframe></span><br></p>

Cutkosky applies analyses, simulations, and experiments to the design and control of robotic hands, tactile sensors, and devices for human/computer interaction. In manufacturing, his work focuses on design tools for rapid prototyping.

robotic hands

ORGANIZATIONS.

SHAPING THE INDUSTRY.

Soft Robotics Podcast

Shadow Robot

Latest Posts