<p><span data-contrast="auto" id="isPasted" lang="EN-US">Non-destructive testing (NDT) of water tanks is a challenging task, as conventional inspection methods, such as rope access or scaffolding, can be time-consuming, expensive, and risky. Water tanks are often large and complex structures with curved and angled surfaces, making it difficult for inspectors to access many areas. Additionally, conventional&nbsp;<iframe id="6433ae788de910000801b259" width="250" class="specs-iframe" height="420" src="https://wevolver.com/iframe/specs/skyron?iframe=true" frameborder="0" scrolling="true" data-lf-form-tracking-inspected-kn9eq4roawkarlvp="true" data-lf-yt-playback-inspected-kn9eq4roawkarlvp="true" data-lf-vimeo-playback-inspected-kn9eq4roawkarlvp="true"></iframe>methods can be obstructive, causing disturbances and limited mobility, making it harder to get a complete view of the tank&#39;s condition. To overcome these challenges, intelligent flying robots offer a more efficient, cost-effective, and safer alternative for water tank inspections.&nbsp;</span><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></p><p><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'><span data-contrast="auto" id="isPasted" lang="EN-US">Using an intelligent flying robot for NDT inspections of water tanks helps to access hard-to-reach areas, providing an unobstructed view of the entire structure for inspectors, significantly reducing the cost and risks associated with conventional inspection methods.&nbsp;Although challenges exist, careful planning and operation can overcome obstacles to ensure that the inspection process is successful, allowing for faster detection of potential damage or corrosion using advanced sensors and providing real-time data for analysis.</span><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></span></p><p><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'><span data-ccp-props="{&quot;201341983&quot;:0,&quot;335559739&quot;:160,&quot;335559740&quot;:259}"></span></span></p><p><br></p><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjgzMjIwMTYwNDMwLVdhdGVyIFRhbmsgMDEtMi5qcGciLCJlZGl0cyI6eyJyZXNpemUiOnsid2lkdGgiOjk1MCwiZml0IjoiY292ZXIifX19" style="width: 300px;" class="fr-fic fr-dib"><p></p><p><span data-contrast="auto" id="isPasted" lang="EN-US">The unique shape of the water tank poses a challenge for the inspection by blocking the line of sight. However, SKYRON overcomes this challenge with its HD camera system that provides a near-zero latency HD video feed to operators, enabling operations beyond line of sight. The First&nbsp;Person&nbsp;View (FPV)&nbsp;camera ensures a clear view of what&#39;s ahead of SKYRON for precise and efficient inspections. Additionally, SKYRON&#39;s strong wind handling capabilities aid in overcoming the challenge of high winds during flight, ensuring successful inspections. During this inspection, winds reached 30 km/h, yet SKYRON&#39;s advanced technology facilitated a successful two-man operation, accurately inspecting the tank shell and legs while improving safety and time management.&nbsp;</span><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></p><p><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjgzMjIwMTk0OTgyLVdhdGVyIFRhbmsgMDIuanBnIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" style="width: 662px;" class="fr-fic fr-dib"></span></p><p><br></p><p id="isPasted"><span data-contrast="auto" lang="EN-US">The intelligent flying&nbsp;not only overcame one obstacle but also tackled another challenge by adjusting its Rotary Arm to match the different angles and complex geometry of the water tank during the inspection. Specifically, the Rotary Arm can be attached to flat, angled, or vertical surfaces by rotating its orientation to mimic the contour of the attachment point.&nbsp;After being attached to the surface, the intelligent flying robot applies&nbsp;couplant&nbsp;to ensure unobstructed measurements. It then uses an integrated gauge to perform ultrasonic thickness (UT) measurements. The robot captures the data and transfers it through the Smart Tether System, which automatically generates a report&nbsp;containing&nbsp;thickness data and an overview of the inspection, including A scans.&nbsp;The&nbsp;results&nbsp;demonstrate&nbsp;the robot&#39;s high efficiency and accuracy, with 50 UT readings taken in only 20 minutes of flight time.</span><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></p><p><span data-contrast="auto" lang="EN-US">Implementing this advanced technology of flying robots in inspecting water tanks&nbsp;represents&nbsp;an efficient and cost-effective solution for the unique challenges presented by the size and geometry of these assets. The success of this inspection highlights the potential for this technology to be&nbsp;utilized&nbsp;in inspecting other assets, leading to a safer and more efficient inspection process.</span><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></p>

Using an intelligent flying robot for NDT inspections of water tanks helps to access hard-to-reach areas

Non-Destructive Testing (NDT) Inspection of Water Tanks Using Intelligent Flying Robots

<p><iframe id="6433ae788de910000801b259" width="250" class="specs-iframe" height="420" src="https://wevolver.com/iframe/specs/skyron?iframe=true" frameborder="0" scrolling="true" data-lf-form-tracking-inspected-kn9eq4roawkarlvp="true" data-lf-yt-playback-inspected-kn9eq4roawkarlvp="true" data-lf-vimeo-playback-inspected-kn9eq4roawkarlvp="true"></iframe></p><p id="isPasted"><span data-contrast="none" lang="EN-US">SKYRON, an intelligent flying robot, can perform Non-Destructive Testing inspections on various assets across various industries, including tanks, towers, pipes, flare stacks, ships, and more.&nbsp;It has been used onshore and offshore and&nbsp;can operate&nbsp;in confined spaces.</span><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></p><p><span data-contrast="none" lang="EN-US">The robot&#39;s exceptional capabilities have been&nbsp;demonstrated&nbsp;through successful contact-based inspections conducted for different companies and service providers over the past few years. As a tool with an unparalleled ability to perform NDT inspections, SKYRON has proven invaluable in various industries.</span><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span><br><br><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'><span data-contrast="none" id="isPasted" lang="EN-US">To see this flying robot in action, watch the video above showcasing its capabilities on different assets and locations.</span><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></span></p>

This flying robot attaches to various assets to perform contact based Non-Destructive Testing

Intelligent flying robots can perform Non-Destructive Testing inspections on various assets across various industries, including tanks, towers, pipes, flare stacks, ships, and more.

FLYING ROBOT'S RANGE OF ASSETS

<p dir="ltr" id="isPasted">Delta arms have become an essential part of modern robotics and automation. These versatile machines are at the forefront of innovation, driving efficiency and precision in numerous industries. In this comprehensive guide, we will explore the fascinating world of delta arms, from their basic structure and components to their wide-ranging applications and future developments.</p><h2 dir="ltr">What is a Delta Arm?</h2><p dir="ltr">A delta arm, also known as a delta robot, is a type of parallel robot that features a unique, triangular base and lightweight, interconnected arms. These arms can move in all three dimensions and are attached to a central end effector, which can be customized for various applications, such as gripping, cutting, or dispensing materials. Delta arms are known for their speed, precision, and flexibility, making them highly sought-after in industries like manufacturing, packaging, and assembly.</p><h2 dir="ltr">History of Delta Arms</h2><p dir="ltr">The delta arm was first invented in the early 1980s by Swiss engineer Reymond Clavel, who was inspired by the desire to create a high-speed, accurate robot for use in the watchmaking industry. Clavel&#39;s invention sparked a revolution in robotics, with companies and researchers around the world quickly adopting and refining delta arm technology. Notable companies like ABB, FANUC, and KUKA participated in the development of delta arms. Today, delta arms are used in a wide array of applications, from industrial automation to 3D printing and even medical procedures.</p><h2 dir="ltr">How Delta Arms Work?</h2><p dir="ltr">At the heart of the delta arm&#39;s impressive performance are its unique design and kinematics. Delta arms employ a parallel mechanism, which allows the arms to move simultaneously and in harmony with one another. This setup enables the end effector to maintain a consistent orientation while moving through its workspace, resulting in highly precise and accurate motion.</p><p dir="ltr">To better understand delta arm kinematics, it&#39;s helpful to visualize the system as a series of interconnected links and joints. As the arms move, the joints rotate, causing the end effector to translate through space. By carefully controlling these rotations, the delta arm can achieve a wide range of positions and orientations with remarkable speed and precision.</p><p dir="ltr"><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/3-NeyKcyfC4?&ab_channel=ArturWiebe&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><br>A prototype of a delta arm showcases its movements by precisely attaching a bottle cap and then removing it with ease.</p><h2 dir="ltr">Components of Delta Arms</h2><h3 dir="ltr">Base</h3><p dir="ltr">The base of a delta arm serves as the foundation for the entire system, providing stability and support for the arms and end effector. Bases come in various designs, but most commonly feature a triangular or circular shape. The material used in the base&#39;s construction can vary depending on the intended application, with common choices including aluminum, steel, and even 3D-printed materials for lighter-weight applications.</p><h3 dir="ltr">Arms</h3><p dir="ltr">The arms of a delta robot are crucial to its movement and overall performance. These slender, interconnected components enable the robot to move with great speed and agility. Arms can be constructed from a variety of materials, such as carbon fiber, aluminum, or steel, with the choice of material typically dependent on the required strength, weight, and cost considerations for the specific application.</p><h3 dir="ltr">Joints</h3><p dir="ltr">Joints are essential to the delta robot kinematics, providing the flexibility and range of motion necessary for the robot to function effectively in its working envelope while exhibiting good repeatability. There are several types of joints used in delta arms, including revolute, prismatic, and ball-and-socket joints. The selection of the appropriate joint type depends on the desired movement and performance characteristics for the specific delta arm application.</p><p dir="ltr"><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/ih3oXigeY-U?&ab_channel=KeyanGhazi-Zahedi&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>A prismatic joint</p><h3 dir="ltr">End effector</h3><p dir="ltr">The end effector is the business end of a delta arm, responsible for carrying out the tasks the robot is designed to perform. End effectors can include grippers, cutters, welding tools, dispensers, vision systems, and more. Some delta arms even feature interchangeable end effectors, allowing them to be easily adapted for different tasks as needed.</p><p dir="ltr"><span class="fr-img-caption fr-fic fr-dii" style="width: 758px;"><span class="fr-img-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjgwNTQyNjMzMTMwLTE2ODA1NDI2MzMxMzAucG5nIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" width="720" height="540" class="fr-fic fr-dii"><span class="fr-inner" spellcheck="false">Fig. 1: A delta arm robot with a customizable end effector</span></span></span></p><p dir="ltr"><strong><em>Suggested reading:&nbsp;</em></strong><a href="https://www.wevolver.com/article/what-are-end-effectors-in-robotics-types-of-end-effectors-applications-future/" rel="noopener noreferrer" target="_blank"><strong><em>What are End Effectors in Robotics? Types of End Effectors, Applications, Future</em></strong></a></p><h3 dir="ltr">Actuators</h3><p dir="ltr">Actuators are the driving force behind a delta arm&#39;s movement, converting electrical or pneumatic energy into mechanical motion. There are several types of actuators used in delta arms, including stepper motors, servo motors, and pneumatic cylinders. The choice of actuator depends on factors such as the required speed, precision, and force, as well as cost considerations for the specific application.</p><p dir="ltr"><strong><em>Suggested reading:&nbsp;</em></strong><a href="https://www.wevolver.com/article/what-is-an-actuator-principles-classification-and-applications" rel="noopener noreferrer" target="_blank"><strong><em>What is an Actuator? Principles, Classification, and Applications</em></strong></a></p><h3 dir="ltr">Controllers</h3><p dir="ltr">Controllers play a crucial role in delta arm operation, managing the movement and coordination of the robot&#39;s various components. Controllers can range from simple microcontrollers to more advanced robot control systems incorporating artificial intelligence and machine learning capabilities. The selection of a controller depends on factors such as the complexity of the delta arm&#39;s movements, the desired level of precision, and the integration requirements with other systems or devices.</p><h2 dir="ltr">Applications of Delta Arms</h2><h3 dir="ltr">Industrial Automation</h3><p dir="ltr">Delta arms have become a staple in production line automation, with their speed, precision, and flexibility making them ideal for applications such as manufacturing, packaging, material handling,&nbsp;and assembly. In manufacturing, delta arms are used to pick and place components, assemble products, and even perform quality control checks. In packaging, they can quickly and accurately fill containers, apply labels, and seal packages. Additionally, delta arms have found a home in assembly lines, where their speed and dexterity make them perfect for tasks such as electronic component placement and assembly.</p><p dir="ltr"><em><strong>Suggested Reading:&nbsp;</strong></em><a href="https://www.wevolver.com/article/what-are-robotic-assembly-lines-history-components-advantages-limitations-applications-and-future" target="_blank"><strong><em>What are Robotic Assembly Lines? History, Components, Advantages, Limitations, Applications, and Future</em></strong></a></p><h3 dir="ltr">3D Printing</h3><p dir="ltr">Delta arms have made a significant impact on the world of 3D printing, with their unique design and movement capabilities making them well-suited for this application. Delta arm-based 3D printers can achieve high levels of accuracy and resolution, while their lightweight arms allow for rapid movement and reduced print times.</p><h3 dir="ltr">Medical and Pharmaceutical Industries</h3><p dir="ltr">Delta arms are increasingly finding applications in the medical and pharmaceutical fields, with their precision and dexterity making them ideal for tasks such as drug discovery, surgery, and patient care. In drug discovery, delta arms can be used to quickly and accurately handle and analyze samples, speeding up the research process.&nbsp;</p><p dir="ltr">In surgery, delta arms have been employed in procedures such as laparoscopy and microsurgery, where their precision and minimal invasiveness offer significant benefits.&nbsp;</p><p dir="ltr">In patient care, delta arms can be used for tasks such as administering medications, drawing blood samples, and even assisting with rehabilitation exercises.</p><h3 dir="ltr">Food and Beverage Industry</h3><p dir="ltr">The food industry has also embraced delta arms for their speed, accuracy, and hygienic properties. Delta arms are used in various aspects of food processing, from sorting and grading fruits and vegetables to handling delicate baked goods. In packaging, delta arms can rapidly and accurately fill containers, apply labels, and seal packages, ensuring product quality and safety. Additionally, delta arms have been utilized in quality control tasks, such as inspecting food items for defects or contamination.</p><h3 dir="ltr">Other Applications</h3><p dir="ltr">While the aforementioned industries represent some of the most common applications for delta arms, these versatile robots have also found their way into many other fields. For example, delta arms have been used in research and development, aerospace, and entertainment industries, to name a few. As technology advances and delta arms become even more adaptable and efficient, their potential applications will undoubtedly continue to grow.</p><h2 dir="ltr">Choosing and Implementing Delta Arms</h2><p dir="ltr">When selecting a delta arm for a specific application, there are several factors to consider, such as payload capacity, degrees of freedom, speed, accuracy, and cost. Payload capacity refers to the maximum weight the delta arm can carry or manipulate, and it is essential to choose a delta arm with the appropriate capacity for the intended task. Degrees of freedom is nothing but the number of joints present in the robot that allow it to move in the x, y, and z-axis. Speed and accuracy are also vital considerations, as they will impact the overall performance and efficiency of the delta arm in its specific application. Finally, cost is always a factor, with the selection of materials, actuators, and controllers playing a significant role in determining the overall price of the delta arm.</p><p dir="ltr"><em><strong>Suggested reading:&nbsp;</strong></em><a href="https://www.wevolver.com/article/7-types-of-industrial-robots-advantages-disadvantages-applications-and-more" target="_blank"><strong><em>7 Types of Industrial Robots: Advantages, Disadvantages, Applications, and More</em></strong></a></p><h3 dir="ltr">Integration with Existing Systems</h3><p dir="ltr">Integrating a delta arm into an existing system can be a complex process, requiring careful planning and consideration. Factors such as compatibility with existing hardware and software, workspace constraints, and safety protocols must all be taken into account. Additionally, training employees to operate and maintain the delta arm is another critical aspect of successful integration.</p><h3 dir="ltr">Maintenance and Troubleshooting</h3><p dir="ltr">Regular maintenance is essential to keep a delta arm functioning optimally and ensure a long service life. This maintenance may include tasks such as lubricating joints, checking for wear and tear on components, and calibrating the system for optimal performance. Familiarity with common troubleshooting techniques is also crucial, as it allows operators to quickly diagnose and resolve issues that may arise during operation. Potential issues can include mechanical failures, software glitches, and problems with the end effector or other components.</p><h2 dir="ltr">The Future of Delta Arms</h2><p dir="ltr">Recent advancements in delta arm technology have focused on improving performance, efficiency, and versatility. Innovations such as lightweight materials, advanced actuators, and more sophisticated controllers have all contributed to the ongoing evolution of delta arms. As technology continues to advance, we can expect even more impressive capabilities from these remarkable machines.</p><p dir="ltr"><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/yXNbG4P8fTU?&t=37s&ab_channel=ABBRobotics&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 dir="ltr">The ABB FlexPicker: a high-throughput variant of a delta robot arm suitable for rapid pick and place applications and more</p><h3 dir="ltr">The Role of AI in Delta Arms</h3><p dir="ltr">Artificial intelligence (AI) is increasingly being integrated with delta arms, allowing for enhanced performance and adaptability. AI-powered delta arms can learn from experience, adapting their movements and strategies to optimize efficiency and effectiveness. As AI continues to advance, we can expect even more exciting developments in the world of delta arm robotics.</p><h3 dir="ltr">The Impact of Delta Arms on the Workforce</h3><p dir="ltr">The rise of delta arms in various industries has raised questions about their potential impact on the workforce and job market. While it is true that delta arms can automate many tasks that humans once performed, they can also complement human labor rather than replace it entirely. By taking over repetitive, mundane, or dangerous tasks, delta arms can free up human workers to focus on more creative, strategic, and higher-value activities. Nevertheless, the integration of delta arms in the workforce also presents challenges, such as the need for retraining and upskilling workers to adapt to new technologies and roles. Addressing these challenges through effective education and training programs will be essential for ensuring a smooth transition and maximizing the benefits of delta arm technology.</p><h2 dir="ltr">Conclusion</h2><p dir="ltr">In conclusion, delta arms have made a significant impact on the world of robotics and automation, offering speed, precision, and flexibility that make them invaluable in a wide range of industries. From manufacturing and packaging to medical procedures and 3D printing, these versatile robots are revolutionizing the way we work and live. As technology continues to advance, we can expect even more exciting developments in the world of delta arms, making them an essential part of our future.</p><h2 dir="ltr">Frequently Asked Questions (FAQs)</h2><ol><li dir="ltr" style="font-weight: bold;"><p dir="ltr"><strong>What are the main benefits of using delta arms in automation?</strong></p></li></ol><p dir="ltr">Delta arms offer numerous advantages, such as speed, precision, flexibility, and cost-effectiveness, making them ideal for a wide range of automation tasks.</p><ol start="2"><li dir="ltr" style="font-weight: bold;"><p dir="ltr"><strong>How do delta arms compare to other robotic arms in terms of performance?</strong></p></li></ol><p dir="ltr">Delta arms are known for their unique combination of speed, precision, and agility, which sets them apart from other types of robotic arms.</p><ol start="3"><li dir="ltr" style="font-weight: bold;"><p dir="ltr"><strong>Can delta arms be used in hazardous environments?</strong></p></li></ol><p dir="ltr">With appropriate modifications and precautions, delta arms can be used in hazardous environments, such as those involving high temperatures, radiation, or corrosive substances.</p><ol start="4"><li dir="ltr" style="font-weight: bold;"><p dir="ltr"><strong>What are the main challenges in implementing delta arms in existing systems?</strong></p></li></ol><p dir="ltr">Potential challenges include integration with existing hardware and software, workspace constraints, safety protocols, and training employees to operate and maintain the delta arm.</p><ol start="5"><li dir="ltr" style="font-weight: bold;"><p dir="ltr"><strong>How much do delta arms typically cost?</strong></p></li></ol><p dir="ltr">The cost of a delta arm can vary widely, depending on factors such as size, capabilities, and materials used. Prices can range from a few thousand dollars for basic models to tens or even hundreds of thousands of dollars for more advanced systems.</p><ol start="6"><li dir="ltr" style="font-weight: bold;"><p dir="ltr"><strong>What is the expected lifespan of a delta arm?</strong></p></li></ol><p dir="ltr">The average lifespan of a delta arm depends on factors such as usage, maintenance, and build quality. With proper care, a delta arm can typically last for many years or even decades.</p><ol start="7"><li dir="ltr" style="font-weight: bold;"><p dir="ltr"><strong>Are delta arms safe to use alongside human workers?</strong></p></li></ol><p dir="ltr">With appropriate safety measures and best practices in place, delta arms can be safely used in collaborative work environments, complementing human labor and improving overall efficiency.</p><h2 dir="ltr">References</h2><p>[1] Delta robot. (2023, March 7). In Wikipedia. <a data-fr-linked="true" href="https://en.wikipedia.org/wiki/Delta_robot" id="isPasted" rel="noopener noreferrer" target="_blank">https://en.wikipedia.org/wiki/Delta_robot</a></p><p>[2] Pierrot F, Reynaud C, Fournier A. DELTA: a simple and efficient parallel robot. Robotica. Cambridge University Press; 1990;8(2):105&ndash;9.&nbsp;</p>

Delta arms are high-performance robotic arms that use parallel linkages and jointed construction to provide fast, precise, and flexible movement, making them ideal for various industrial and manufacturing applications.

The Ultimate Guide to Delta Arms: Revolutionizing Robotics and Automation

<p id="isPasted"><span data-contrast="none" lang="EN-US">Case Study: SKYRON Inspects Three Assets in a Refinery</span><span data-ccp-props='{"134233117":false,"134233118":false,"201341983":0,"335551550":1,"335551620":1,"335559685":0,"335559737":0,"335559738":0,"335559739":160,"335559740":259}'>&nbsp;</span></p><p><span data-contrast="none" lang="EN-US">Refineries are known for operating under harsh conditions of high pressure and temperature, which can cause damage to equipment over time, leading to defects such as internal corrosion, cracking, and erosion. Non-Destructive Testing (NDT) inspections are necessary for refineries to ensure safety and reliability. These inspections can detect any defects in pipes, tanks, flare stacks, and other equipment that may affect their structural integrity and functionality. Regular&nbsp;Ultrasonic&nbsp;Thickness (UT)&nbsp;inspections help refineries maintain safe&nbsp;and&nbsp;reliable&nbsp;equipment, which prevents incidents, reduces downtime, and increases productivity.</span><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></p><p><span data-contrast="none" lang="EN-US">Conventional inspection methods in refineries are known to be costly, time-consuming, and pose risks to inspectors. However, technological advancements have led to the development of intelligent flying robots that can perform inspections faster, cost-effectively, and safely. These&nbsp;robots can avoid downtime as inspections can be done while the equipment is in service, resulting in increased productivity and cost savings.</span><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></p><p><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjgwMTI0MTA2MTc1LTE2ODAxMjQxMDYxNzUuanBlZyIsImVkaXRzIjp7InJlc2l6ZSI6eyJ3aWR0aCI6OTUwLCJmaXQiOiJjb3ZlciJ9fX0=" class="fr-fic fr-dii"><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></p><p><span data-contrast="none" lang="EN-US">&nbsp;<iframe id="6433ae788de910000801b259" width="250" class="specs-iframe" height="420" src="https://wevolver.com/iframe/specs/skyron?iframe=true" frameborder="0" scrolling="true" data-lf-form-tracking-inspected-kn9eq4roawkarlvp="true" data-lf-yt-playback-inspected-kn9eq4roawkarlvp="true" data-lf-vimeo-playback-inspected-kn9eq4roawkarlvp="true"></iframe>Recently, the intelligent flying robot&nbsp;SKYRON&nbsp;took&nbsp;UT measurements on three assets:&nbsp;a tank, a complex network of pipes, and an in-service flare stack. The inspection presented several challenges due to the&nbsp;</span><span data-contrast="none" lang="EN-US">height</span><span data-contrast="none" lang="EN-US">&nbsp;of the tank shell and roof,&nbsp;the&nbsp;</span><span data-contrast="none" lang="EN-US">requirement</span><span data-contrast="none" lang="EN-US">&nbsp;to inspect</span><span data-contrast="none" lang="EN-US">&nbsp;</span><span data-contrast="none" lang="EN-US">pipes</span><span data-contrast="none" lang="EN-US">&nbsp;from</span><span data-contrast="none" lang="EN-US">&nbsp;</span><span data-contrast="none" lang="EN-US">various angles</span><span data-contrast="none" lang="EN-US">, and th</span><span data-contrast="none" lang="EN-US">e&nbsp;</span><span data-contrast="none" lang="EN-US">high temperatures</span><span data-contrast="none" lang="EN-US">&nbsp;of&nbsp;the flare stack.</span><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></p><p><span data-contrast="none" lang="EN-US">The inspection process in the refinery began with an examination of a tank, which is challenging for inspectors due to the risks associated with working at elevated heights and encountering obstructions or limited access points. Fortunately, an intelligent flying robot has simplified this process and allowed inspectors to operate securely on the ground. With its advanced HD camera system, SKYRON enables operators to conduct inspections beyond the line of sight and navigate to limited access points, such as the tank roof, even when the robot is out of sight due to distance. Furthermore, SKYRON&#39;s wind handling capabilities and&nbsp;100-meter&nbsp;tether cable allow operators to take UT measurements from the tank shell and roof&nbsp;with ease and precision, reducing the risks of accidents associated with high winds and obstructions. SKYRON enabled inspectors to take 155 readings within just 45 minutes of flight time, eliminating the need for high-altitude climbing and significantly reducing the risks involved for inspectors.</span><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></p><p><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjgwMTI0MTA2NTMyLTE2ODAxMjQxMDY1MzIucG5nIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" class="fr-fic fr-dii"><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></p><p><span data-contrast="none" lang="EN-US">The second asset on the inspection list was a complex network of&nbsp;pipes. Inspecting these structures comprehensively requires examining them from multiple angles, which can be challenging and may require disassembling the pipes and elbows, causing disruptions and delays in refinery operations. Additionally, these structures may be located in hard-to-reach areas or obscured by other equipment, further complicating the inspection process.&nbsp;Intelligent flying robots have revolutionized the inspection process by allowing easy access to difficult-to-reach areas and providing multiple inspection angles, even while the pipes are still in service. This has greatly improved efficiency and reduced downtime during inspections. The intelligent flying robot SKYRON has a Rotary Arm that can adapt to any surface slope, providing measurements within a 2&quot; x 2&quot; square. The arm pivots 180 degrees to accommodate any gradient, regardless of orientation, ensuring that even the most challenging pipe surfaces can be inspected thoroughly.&nbsp;During the inspection, thanks to its advanced technology, SKYRON took 94 UT readings in 35 minutes of flight.</span><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></p><p><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjgwMTI0MTA2NjM5LTE2ODAxMjQxMDY2MzgucG5nIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" class="fr-fic fr-dii"><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></p><p><span data-contrast="none" lang="EN-US">The third asset in question was an in-service flare stack, which posed a significant challenge to inspect due to its towering height and extreme temperature conditions.&nbsp;Flare stacks require shutting down&nbsp;operations&nbsp;to reduce the heat&nbsp;and&nbsp;make the conditions safe for inspectors,&nbsp;resulting in reduced productivity.&nbsp;By utilizing intelligent flying robots for inspections, the need for shutdowns can be eliminated. The robot remains in contact with the surface being inspected, while inspectors can maintain a safe distance, ensuring maximum efficiency and safety.&nbsp;The advanced technology of&nbsp;SKYRON&nbsp;allowed it to navigate the high-temperature&nbsp;surface and took 27 readings in 8 minutes of flight time.</span><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></p><p><span data-contrast="none" lang="EN-US">Using flying robots for NDT inspections proves that technological advancements&nbsp;can improve efficiency and safety in the&nbsp;oil&nbsp;and gas&nbsp;industry. With the ability to perform inspections faster, cost-effectively, and safely, intelligent flying robots like SKYRON can reduce downtime and increase productivity, saving refineries time and money. As the industry continues to evolve, it is clear that the use of advanced technology will play a vital role in ensuring the safety and reliability of equipment.</span><span data-ccp-props='{"201341983":0,"335559739":160,"335559740":259}'>&nbsp;</span></p>

Refineries are known for operating under harsh conditions of high pressure and temperature, which can cause damage to equipment over time, leading to defects such as internal corrosion, cracking, and erosion. 

Flying Robot Inspects Tanks, Pipes, and Flare stacks

<p dir="ltr" id="isPasted">Articulated robots&nbsp;are highly versatile&nbsp;industrial robots&nbsp;that have been used in various industries to perform a wide range of tasks. These robots are designed with a series of interconnected segments, known as links, which are attached through movable joints. This design allows them to move with a high degree of flexibility and dexterity, making them ideal for performing complex tasks requiring high precision.</p><p dir="ltr">In this article, we will delve deeper into the specifics of&nbsp;articulated robots, explore the various&nbsp;control systems&nbsp;that they use, discuss their advantages and limitations, and examine their applications across different industries. We will also look at the challenges of operating&nbsp;articulated robots&nbsp;and discuss future trends and developments in this field.</p><h1 dir="ltr">Introduction</h1><p dir="ltr">The history of&nbsp;articulated robots&nbsp;dates back to the 1950s, when they were first introduced for use in&nbsp;manufacturing processes. Since then, they have evolved significantly and become integral to many industries, including&nbsp;automotive, aerospace, healthcare, and research.</p><p dir="ltr">The importance of&nbsp;articulated robots&nbsp;lies in their ability to improve efficiency, accuracy, and safety in various operations. These robots are designed to perform repetitive and dangerous tasks that are often too difficult or hazardous for humans to undertake. As a result, they have helped to increase productivity and reduce the risk of workplace accidents.</p><p dir="ltr"><em><strong>Suggested Reading:</strong></em><a href="https://www.wevolver.com/article/7-types-of-industrial-robots-advantages-disadvantages-applications-and-more" target="_blank"><strong><em>&nbsp;7 Types of Industrial Robots: Advantages, Disadvantages, Applications, and More</em></strong></a></p><h1 dir="ltr">Anatomy of Articulated Robots</h1><p dir="ltr">Articulated robots&nbsp;comprise several interconnected segments called links, which are connected through joints. These joints are designed to allow the robot to move with a high degree of flexibility, precision, and dexterity. The links of&nbsp;articulated robots&nbsp;can range from two to six or more, with each link providing a degree of freedom (DOF) that enables the robot to move in various directions.</p><p dir="ltr">The joints used in&nbsp;articulated robots&nbsp;can be classified into two types: revolute and prismatic. Revolute joints are&nbsp;rotary joints&nbsp;that allow the robot to rotate along an axis. Prismatic joints, on the other hand, are linear joints that enable the robot to move in a straight line.</p><p dir="ltr">Articulated robots&nbsp;are designed to have multiple&nbsp;degrees of freedom&nbsp;(DOF), which is the number of independent parameters that determine the robot&#39;s position and orientation. This allows the robot to move in a wide range of directions and positions, making it highly versatile. The DOF of&nbsp;articulated robots&nbsp;can range from two to six or more, with each DOF representing a degree of movement or rotation.</p><p dir="ltr">The most common DOF configurations for&nbsp;articulated robots&nbsp;are 4-DOF and 6-DOF. 4-DOF robots typically have three revolute joints and one prismatic joint, while 6-DOF robots have three revolute joints and three prismatic joints. The additional&nbsp;degrees of freedom&nbsp;in 6-DOF robots allow for greater flexibility and precision in their movements.</p><p dir="ltr">In addition to the links and joints,&nbsp;articulated robots&nbsp;also have&nbsp;end-effectors, which are the devices attached to the&nbsp;articulated&nbsp;robot arm&nbsp;that allow it to perform specific tasks.&nbsp;End-effectors&nbsp;can include grippers, suction cups, and other specialized tools that are designed to manipulate objects with precision and accuracy.</p><p dir="ltr"><em><strong>Suggested Reading:</strong></em><a href="https://www.wevolver.com/article/what-are-end-effectors-in-robotics-types-of-end-effectors-applications-future" target="_blank"><strong><em>&nbsp;What are End Effectors in Robotics? Types of End Effectors, Applications, Future</em></strong></a></p><h1 dir="ltr">Control Systems of Articulated Robots</h1><p dir="ltr">The&nbsp;control systems&nbsp;of&nbsp;articulated robots&nbsp;are crucial in determining the robot&#39;s movements, actions, and overall performance. The&nbsp;robot control systems&nbsp;are responsible for monitoring the robot&#39;s position, speed, and orientation and making adjustments as needed to ensure that the robot moves accurately and precisely.</p><p dir="ltr">There are two main types of&nbsp;control systems&nbsp;used in&nbsp;articulated robots: open-loop and closed-loop&nbsp;control systems. Open-loop&nbsp;control systems&nbsp;are used in simple, repetitive tasks where the robot&#39;s movement can be pre-programmed and does not require real-time feedback. Closed-loop&nbsp;control systems, on the other hand, use sensors to monitor the robot&#39;s movements and make adjustments in real-time to ensure precise and accurate movements.</p><p dir="ltr">Sensors are an essential component of the&nbsp;control systems&nbsp;in&nbsp;articulated robots. They are used to monitor the robot&#39;s movements, track its position and orientation, and detect any errors or deviations from the intended path. Common sensors used in&nbsp;articulated robots&nbsp;include encoders, accelerometers, and proximity sensors.</p><p dir="ltr"><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/HgDEqlhjhrE?&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><br></p><p dir="ltr">A video showcasing&nbsp;articulated robot&nbsp;application&nbsp;on a&nbsp;production line. Notice how the robot displays a wide&nbsp;range of motions&nbsp;in the&nbsp;workspace.</p><p dir="ltr">Actuators are another critical component of the&nbsp;control systems&nbsp;in&nbsp;articulated robots. Actuators are responsible for moving the robot&#39;s joints and links, enabling it to move with precision and accuracy. The most commonly used actuators in&nbsp;articulated robots&nbsp;are electric motors (such as the&nbsp;servo motor), hydraulic cylinders, and pneumatic cylinders.</p><p dir="ltr">The&nbsp;control systems&nbsp;of&nbsp;articulated robots&nbsp;are typically operated through a computer interface. The operator can program the robot&#39;s movements, set its parameters, and monitor its performance through a graphical user interface (GUI). The GUI provides real-time feedback on the robot&#39;s movements, allowing the operator to make adjustments as needed to ensure that the robot moves accurately and precisely.</p><p dir="ltr">In recent years, there have been significant advancements in the&nbsp;control systems&nbsp;of&nbsp;articulated robots, with the integration of artificial intelligence (AI) and machine learning (ML) technologies. These technologies enable the robot to learn and adapt to its environment, improving its performance and increasing its flexibility in performing tasks.</p><h1 dir="ltr">Advantages and Limitations of Articulated Robots</h1><p dir="ltr">Articulated robots&nbsp;offer several advantages over traditional industrial&nbsp;automation&nbsp;methods. Some of these advantages include:</p><ol><li dir="ltr"><p dir="ltr"><strong>Flexibility:</strong> Articulated robots have a high degree of flexibility and dexterity, enabling them to perform a wide range of tasks in various industries.</p></li><li dir="ltr"><p dir="ltr"><strong>Precision:</strong> The use of advanced control systems, sensors, and actuators allows articulated robots to perform tasks with a high degree of precision and accuracy.</p></li><li dir="ltr"><p dir="ltr"><strong>Efficiency:&nbsp;</strong>Articulated robots&nbsp;can perform tasks with greater speed and efficiency than human workers, resulting in increased productivity and reduced labor costs.</p></li><li dir="ltr"><p dir="ltr"><strong>Safety:&nbsp;</strong>Articulated robots&nbsp;can perform tasks that are hazardous or dangerous for human workers, reducing the risk of workplace accidents and injuries.</p></li></ol><p dir="ltr">However, there are also limitations to the use of&nbsp;articulated robots. Some of these limitations include the following:</p><ol><li dir="ltr"><p dir="ltr"><strong>Cost:&nbsp;</strong>Articulated robots&nbsp;can be expensive to purchase and maintain, making them less accessible to small and medium-sized businesses.</p></li><li dir="ltr"><p dir="ltr"><strong>Complex Programming:&nbsp;</strong>Programming&nbsp;articulated robots&nbsp;can be complex and time-consuming, requiring specialized knowledge and expertise.</p></li><li dir="ltr"><p dir="ltr"><strong>Maintenance:&nbsp;</strong>The maintenance of&nbsp;articulated robots&nbsp;can be challenging, requiring specialized skills and equipment.</p></li><li dir="ltr"><p dir="ltr"><strong>Space requirements:&nbsp;</strong>Articulated robots&nbsp;require a significant amount of space to operate, which can be a limitation in smaller facilities.</p></li><li dir="ltr"><p dir="ltr"><strong>Lack of adaptability:&nbsp;</strong>Articulated robots&nbsp;may struggle to perform tasks in constantly changing or unpredictable environments, as they are programmed to perform specific tasks in a set environment.</p></li></ol><p dir="ltr">While&nbsp;articulated robots&nbsp;offer several advantages over traditional industrial&nbsp;automation&nbsp;methods, it is important to consider their limitations and potential challenges before implementing them in a business or manufacturing setting.</p><h1 dir="ltr">Applications of Articulated Robots</h1><p dir="ltr">Articulated robots&nbsp;are used in a wide range of industries and applications, including manufacturing, healthcare, and entertainment. Here are some of the most common applications of&nbsp;articulated robots:</p><ol><li dir="ltr"><p dir="ltr"><strong>Manufacturing:&nbsp;</strong>Articulated robots&nbsp;are widely used in manufacturing industries for tasks such as assembly, welding, painting, packaging, and&nbsp;material handling. These robots can perform these tasks with high precision and speed, resulting in increased productivity and efficiency.</p></li><li dir="ltr"><p dir="ltr"><strong>Healthcare:&nbsp;</strong>Articulated robots&nbsp;are increasingly being used in healthcare applications, such as surgery and rehabilitation. They can perform surgical procedures with greater precision and accuracy than human surgeons, reducing recovery times and improving patient outcomes.</p></li><li dir="ltr"><p dir="ltr"><strong>Entertainment:&nbsp;</strong>Articulated robots&nbsp;are used in the entertainment industry for tasks such as animatronics, special effects, and puppetry. They can be programmed to perform complex movements and actions, resulting in lifelike performances and realistic special effects.</p></li></ol><p dir="ltr"><span class="fr-img-caption fr-fic fr-dii" style="width: 758px;"><span class="fr-img-wrap"><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1679837679185-1679837679185.png" width="720" height="480" id="isPasted" class="fr-fic fr-dii" alt="articulated-robot-arm-manufacturing-application"><span class="fr-inner" spellcheck="false">Fig. 1: Articulated robot used for pick and place application</span></span></span></p><ol start="4"><li dir="ltr"><p dir="ltr"><strong>Agriculture:&nbsp;</strong>Articulated robots&nbsp;are used in agriculture for tasks such as harvesting, planting, and spraying. By operating in difficult and dangerous environments,&nbsp;articulated robots&nbsp;reduce the risk of injury to human workers and improve the efficiency of the agricultural process.</p></li><li dir="ltr"><p dir="ltr"><strong>Military and Defense:&nbsp;</strong>Articulated robots&nbsp;are used in military and defense applications for tasks such as bomb disposal and reconnaissance. These robots can operate in hazardous environments, reducing the risk of injury to military personnel and improving the safety and effectiveness of military operations.</p></li><li dir="ltr"><p dir="ltr"><strong>Education:&nbsp;</strong>Articulated robots&nbsp;are increasingly being used in educational settings to teach students about robots and&nbsp;robotic automation. Robots are specifically programmed to perform a specific set of tasks, allowing students to learn about robotics in a hands-on, interactive way.</p></li></ol><h1 dir="ltr">Selection Criteria for Articulated Robots</h1><p dir="ltr">When it comes to selecting an&nbsp;industrial robot, there are a variety of factors to consider, including the application,&nbsp;payload&nbsp;capacity, reach, and&nbsp;repeatability&nbsp;required. In some cases, an&nbsp;articulated robot&nbsp;may be the best choice, while in others, a different&nbsp;type of robot, such as a&nbsp;Cartesian robot, a&nbsp;SCARA robot, or a&nbsp;Delta robot, may be more appropriate.</p><p dir="ltr"><em><strong>Suggested Reading:</strong></em><a href="https://www.wevolver.com/article/what-are-robot-gantries-types-applications-advantages-selection-criteria-and-more" target="_blank"><strong><em>&nbsp;What are Robot Gantries? Types, Applications, Advantages, Selection Criteria, and more</em></strong></a></p><p dir="ltr">Articulated robots&nbsp;are often the preferred choice when a high degree of flexibility and maneuverability is required. These robots have multiple joints that can move independently, allowing them to reach a wide range of positions and orientations.</p><p dir="ltr">The ability of&nbsp;articulated robots&nbsp;to move in multiple directions makes them well-suited for applications that require complex movements. They are also capable of handling a variety of&nbsp;payloads, from small electronic components to heavy&nbsp;automotive&nbsp;parts.</p><p dir="ltr">In contrast,&nbsp;Cartesian robots&nbsp;are designed to move in a straight line along three axes, making them ideal for applications that require precise linear motion in a cubical&nbsp;work envelope. Another type of&nbsp;industrial robot&nbsp;is the&nbsp;SCARA robot, which is designed for tasks that require&nbsp;high-speed,&nbsp;precise movement&nbsp;in a horizontal plane.</p><p dir="ltr">When deciding whether to use an&nbsp;articulated robot&nbsp;or another type of&nbsp;industrial robot, it is important to consider the specific requirements of the application.&nbsp;Articulated robots&nbsp;are best suited for applications that require a high degree of flexibility and maneuverability, while Cartesian and&nbsp;SCARA robots&nbsp;are better suited for applications that require precise linear or horizontal movement.</p><p dir="ltr">Ultimately, the selection of an&nbsp;industrial robot&nbsp;will depend on the application&#39;s specific needs, as well as factors such as cost, ease of programming, and maintenance requirements. By carefully evaluating the application&#39;s requirements, it is possible to choose the right&nbsp;type of robot&nbsp;for the job and ensure maximum efficiency and productivity.</p><h1 dir="ltr">Future Developments in Articulated Robots</h1><p dir="ltr">As technology continues to evolve,&nbsp;articulated robots&nbsp;are expected to become even more advanced and versatile. Here are some of the developments that we can expect to see in the future of&nbsp;articulated robotics:</p><ol><li dir="ltr"><p dir="ltr"><strong>Enhanced sensory capabilities:&nbsp;</strong>As sensor technology continues to improve,&nbsp;articulated robots&nbsp;will be able to sense and respond to their environment in more sophisticated ways. This will allow robots to adapt to changing environments and perform more complex tasks.</p></li><li dir="ltr"><p dir="ltr"><strong>Improved mobility:&nbsp;</strong>Advances in mobility technology, such as the development of new types of actuators and locomotion systems, will allow&nbsp;articulated robots&nbsp;to move more efficiently and effectively in a wider range of environments.</p></li><li dir="ltr"><p dir="ltr"><strong>Artificial intelligence:&nbsp;</strong>The integration of artificial intelligence and machine learning algorithms into&nbsp;articulated robots&nbsp;will enable them to learn and adapt to new tasks and environments more quickly and efficiently.</p></li><li dir="ltr"><p dir="ltr"><strong>Collaborative robots:&nbsp;</strong>Collaborative robots, or &quot;cobots,&quot; will become more prevalent in the future of&nbsp;articulated robotics. These robots are designed to work alongside human workers, performing tasks that are too dangerous, repetitive, or strenuous for humans to perform.</p></li><li dir="ltr"><p dir="ltr"><strong>Miniaturization:&nbsp;</strong>Advances in miniaturization technology will allow for the development of smaller, more compact&nbsp;articulated robots&nbsp;that can perform tasks in tight spaces or delicate environments.</p></li></ol><h1 dir="ltr">Conclusion</h1><p dir="ltr">Articulated robots&nbsp;have revolutionized the field of industrial&nbsp;automation, offering greater flexibility, precision, and efficiency than traditional&nbsp;factory automation&nbsp;methods. These robots are used in various industries and applications, from manufacturing and healthcare to entertainment and agriculture.</p><p dir="ltr">While there are limitations to the use of&nbsp;articulated robots, such as cost and programming complexity, their advantages make them an attractive option for businesses looking to improve their productivity and efficiency.</p><p dir="ltr">Looking to the future, we can expect to see continued advancements in&nbsp;articulated robot&nbsp;technology, such as enhanced sensory capabilities, improved mobility, and the integration of artificial intelligence and machine learning algorithms. These developments will further increase the&nbsp;versatility&nbsp;and efficiency of&nbsp;articulated robots, allowing them to perform even more complex tasks in a wider range of environments.</p><p dir="ltr">The future of&nbsp;articulated robotics&nbsp;is promising, and businesses that embrace this technology will likely see significant benefits in increased productivity, efficiency, and safety.</p>

Articulated robots are highly versatile and efficient robots that have the ability to mimic human arm movements for various industrial and commercial applications.

What are Articulated Robots? Anatomy, Control Systems, Advantages, Selection Criteria, and Applications

<section id="isPasted"><p>Robots are all around us, from drones filming videos in the sky to serving food in restaurants and diffusing bombs in emergencies. Slowly but surely, robots are improving the quality of human life by augmenting our abilities, freeing up time, and enhancing our personal safety and well-being. While existing robots are becoming more proficient with simple tasks, handling more complex requests will require more development in both mobility and intelligence.&nbsp;</p><p>Columbia Engineering and Toyota Research Institute computer scientists are delving into psychology, physics, and geometry to create algorithms so that robots can adapt to their surroundings and learn how to do things independently. This work is vital to enabling robots to address new challenges stemming from an aging society and provide better support, especially for seniors and people with disabilities.</p></section><section><h2>Teaching robots occlusion and object permanence</h2><p>A longstanding challenge in computer vision is object permanence, a well-known concept in psychology that involves understanding that the existence of an object is separate from whether it is visible at any moment. It is fundamental for robots to understand our ever-changing, dynamic world. But most applications in computer vision ignore occlusions entirely and tend to lose track of objects that become temporarily hidden from view.</p><section id="isPasted"><p>&quot;Some of the hardest problems for artificial intelligence are the easiest for humans,&quot; said <a href="http://www.cs.columbia.edu/~vondrick/" rel="nofollow" target="_blank">Carl Vondrick</a>, associate professor of <a href="https://www.cs.columbia.edu/" rel="nofollow" target="_blank">computer science</a> and a recipient of the Toyota Research Institute Young Faculty award. Think about how toddlers play peek-a-boo and learn that their parent does not disappear when they cover their faces. Computers, on the other hand, lose track once something is blocked or hidden from view and cannot process where the object went or recall its location.&nbsp;</p></section><section><blockquote><p>Some of the hardest problems for artificial intelligence are the easiest for humans</p><footer>CARL VONDRICKASSOCIATE PROFESSOR OF COMPUTER SCIENCE</footer></blockquote></section><section><p>To tackle this issue, Vondrick and his team taught neural networks the basic physical concepts that come naturally to adults and children. Similar to how a child learns physics by watching events unfold in their surroundings, the team created a machine that watches many videos to learn physical concepts. The key idea is to train the computer to anticipate what the scene would look like in the future. By training the machine to solve this task across many examples, the machine automatically creates an internal model of how objects physically move in typical environments. For example, when a soda can disappears from sight inside the refrigerator, the machine learns to remember it still exists because it appears again once the refrigerator door reopens.&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/Fn0XKUhkfqA?&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>Dynamic scene completion: given a monocular video as input, the framework produces a 4D representation that captures the entire scene content along with all the static and dynamic objects within it over time. Credit: Basile Van Hoorick&nbsp;</em></p><section id="isPasted"><p>&quot;I have worked with images and videos before, but getting neural networks to work well with 3D information is surprisingly tricky,&quot; said <a href="https://basile.be/about-me/" rel="nofollow" target="_blank">Basile Van Hoorick</a>, a third-year PhD student who worked with Vondrick to <a data-lf-fd-inspected-kn9eq4roawkarlvp="true" href="https://arxiv.org/pdf/2204.10916.pdf" rel="nofollow" target="_blank">develop the framework that can understand occlusions as they occur</a>. Unlike humans, an understanding of the three-dimensionality of our world does not come naturally to computers. The second leap in the project was not only to convert data from cameras into 3D seamlessly but also to reconstruct the entire configuration of the scene beyond what can be seen.</p><p>This work could expand the perception capabilities of home robots widely. In any indoor environment, things become hidden from view all the time. Hence, robots need to interpret their surroundings intelligently. The &quot;soda can inside the refrigerator&quot; situation is one of many examples. Still, it is easy to see how any application that uses vision will benefit if robots can draw upon their memory and object-permanence reasoning skills to keep track of both objects and humans as they move around the house.</p></section><section><h2>Moving beyond the rigid body assumption</h2><p>Most robots today are programmed with a series of assumptions for them to work. One is the rigid body assumption, which assumes that an object is solid and doesn&#39;t change shape. And that simplifies a lot of things. Roboticists can completely ignore the physics of the object the robot is interacting with and only have to think about the robot&#39;s motion.&nbsp;</p><section id="isPasted"><p>The <a href="https://cair.cs.columbia.edu/" rel="nofollow" target="_blank">Columbia Artificial Intelligence and Robotics (CAIR) Lab</a>, led by Computer Science Assistant Professor Shuran Song, has been researching robotic movement in a different way. Her research focuses on deformable, non-rigid objects--they fold, bend, and change shape. When working with deformable objects, roboticists can no longer rely on the rigid body assumption, forcing them to think about physics again.&nbsp;</p><p>&quot;In our work, we are trying to investigate how humans intuitively do things,&quot; said <a href="https://www.cs.columbia.edu/~shurans/" rel="nofollow" target="_blank">Shuran Song</a>, also a Toyota Research Institute Young Faculty awardee. Instead of trying to account for every possible parameter, her team developed an algorithm that allows the robot to learn from doing, making it more generalizable and lessening the need for massive amounts of training data. It forced the group to rethink how people do an action, like hitting a target with a rope. We usually don&rsquo;t think about the trajectory of the string--instead, we try to hit the object first and adjust our movements until we&rsquo;re successful. &quot;This new perspective was essential to solve this difficult problem in robotics,&rdquo; Song noted.</p></section><section><blockquote><p>In our work, we are trying to investigate how humans intuitively do things.</p><footer>SHURAN SONGASSISTANT PROFESSOR OF COMPUTER SCIENCE</footer></blockquote></section><section><p>Her team won a <a href="https://roboticsconference.org/program/awards/" rel="nofollow" target="_blank">Best Paper Award</a> at the Robot Science and Systems Conference (RSS 2022) for an algorithm they developed, <a href="https://irp.cs.columbia.edu/" rel="nofollow" target="_blank">Iterative Residual Policy (IRP)</a>. IRP is a general learning framework for repeatable tasks with complex dynamics, where a single model was trained using inaccurate simulation data. The algorithm can learn from that data and hit many targets with unfamiliar ropes in robotic experiments, reaching sub-inch accuracy and demonstrating its strong generalization capability.</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/KZNHYY4XjOY?&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>The robot learned how to hit the target (the yellow cup) in seven tries. Credit: Cheng Chi&nbsp;</em></p><p id="isPasted">&quot;Previously, to achieve this level of precision, the robot needed to do the task maybe 100 to 1,000 times,&quot; said <a href="https://cheng-chi.github.io/" rel="nofollow" target="_blank">Cheng Chi</a>, a third-year PhD student who worked with Song to develop IRP. &quot;With our system, we can do it within ten times, which is about the same performance as a person.&quot; &nbsp;</p><p>The researchers noticed that there were still some limitations with the flinging motion that their robot could make. While the flinging motion is effective, it is limited by the speed of the robot arm, which means it cannot handle large items. Not to mention that it is dangerous to have a fast flinging motion around people.&nbsp;</p><p>Song&rsquo;s team took this research a step further and developed a new approach to manipulating them by using actively blown air. They armed their robot with an air pump and it was able to quickly unfold a large piece of cloth or open a plastic bag. <a href="https://dextairity.cs.columbia.edu/" rel="nofollow" target="_blank">The self-supervised learning framework they call DextAIRity</a> learns to effectively perform a target task through a sequence of grasping or air-based blowing actions. Using visual feedback, the system uses a closed-loop formulation that continuously adjusts its blowing direction.</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/iEU7Z29g_M8?&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>DextAIRity&rsquo;s learning-based approach can quickly and reliably open the majority of bags tested. Credit: Zhenjia Xu&nbsp;</em></p><section id="isPasted"><p>&quot;One of the interesting strategies the system developed with the bag-opening task is to point the air a little above the plastic bag to keep the bag open,&quot; said Zhenjia Xu, a fourth-year PhD student who works with Song in the CAIR Lab. &quot;We did not annotate or train it in any way; it learned it by itself.&quot;</p></section><section><h2>What needs to be done to make robots more useful in our homes?</h2><p>Currently, robots can successfully maneuver through a structured environment with clearly defined areas and do one task simultaneously. However, a truly useful home robot should have various skills, be able to work in an unstructured environment, like a living room with toys on the floor, and handle different situations. These robots will also need to know how to identify a task and which subtasks must be done in what order. And then, they will need to know what to do next if they fail at a job and how to adapt to the next steps needed to accomplish their goal.&nbsp;</p><p>&ldquo;The progress that Carl Vondrick and Shuran Song have made with their research contributes directly to Toyota Research Institute&#39;s mission,&quot; says Dr. Eric Krotkov, advisor to the University Research Program. &quot;TRI&#39;s research in robotics and beyond focuses on developing the capabilities and tools to address the socioeconomic challenges of an aging society, labor shortage, and sustainable production. Endowing robots with the capabilities to understand occluded objects and handle deformable objects will enable them to improve the quality of life for all.&rdquo;</p><p>Song and Vondrick plan to collaborate to combine their respective expertise in robotics and computer vision to create robots that assist people in the home. By teaching machines to understand everyday objects in homes, such as clothes, food, and boxes, the technology could enable robots to assist people with mobility disabilities and improve the quality of everyday life for people. By increasing the number of objects and physical concepts that can be learned by robots, the team aims to make these applications possible in the future.</p></section></section></section></section></section>

Columbia computer scientists work with the Toyota Research Institute to make advanced home robots a reality

Engineers Use Psychology, Physics, and Geometry to Make Robots More Intelligent

<p id="isPasted">Coordinated Movement has always been one of the primary challenges to tackle in robotics. Locomotion and manipulation are the two main concepts of movement. Locomotion describes the method a robot uses to transport itself from one place to another. This can take the form of legged, wheeled, or tracked motion. Manipulation, on the other hand, describes the manner in which a robot interacts with objects around it. That includes grabbing an object, picking something up, throwing something, or even opening a door.</p><p>For years, researchers and engineers have studied ways to control locomotion and manipulation simultaneously. Particularly for legged robots, attaching an arm can enable them to perform several mobile manipulation tasks that wheeled or tracked robots cannot perform. The conventional way of controlling such legged manipulators has been to control locomotion and manipulation separately - otherwise known as decoupling mobility and manipulation. While this method has its perks, it does require a lot of engineering efforts to maintain sufficient coordination between the legs and the arm.&nbsp;&nbsp;<iframe id="638d493a4c420f000872b315" width="250" class="specs-iframe" height="400" src="https://wevolver.com/iframe/specs/unitree-robotics-go1-edu-plus?iframe=true" frameborder="0" scrolling="true" data-lf-form-tracking-inspected-kn9eq4roawkarlvp="true" data-lf-yt-playback-inspected-kn9eq4roawkarlvp="true" data-lf-vimeo-playback-inspected-kn9eq4roawkarlvp="true"></iframe></p><p>Researchers refer to modeling an arm-carrying legged robot as &ldquo;a dynamic, high-degree-of-freedom, and non-smooth control problem.&rdquo; This means the complexity of such a model would limit its operation to within specified settings. Venturing into different real-life scenarios would make it increasingly complex and burdensome, and errors can arise, leading to undesirable and unnatural movements.</p><p>This has driven engineers to explore different methods that can bring synergy between the limbs in a way that resembles biologically coordinated movement. Think of the way your arms and legs move and how interconnected they are. For instance, it is difficult to move your left arm and your left leg in different motions while standing. The same applies to your right side. Or consider the activity of picking up something on the ground. Your legs and arm would work together to help you reach the object and pick it up. This is referred to as interlimb neural coupling. In simpler terms, it is called <strong>whole-body control</strong>. It allows for coordination between the limbs and the extension of the capabilities of individual parts.</p><p><br></p><p dir="ltr">In a similar sense, researchers at Carnegie Mellon University have successfully developed a unified policy that governs an arm-carrying legged robot in a whole-body control manner using the concept of reinforcement learning (RL). This allowed them to train the robot to synergize its locomotion and manipulation and perform specific tasks through a dynamic and agile behavior.</p><p dir="ltr"><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/t6NoCx_zitM?autoplay=1&mute=1&wmode=opaque&rel=0" 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><h2 dir="ltr" id="isPasted">A Unified Policy for Coordinated Locomotion and Manipulation</h2><p>Earlier this year, Zipeng Fu, Xuxin Cheng, and Deepak Pathak from Carnegie Mellon University<sup>[1]</sup> set out to test whether a learning-based, unified policy would outperform existing decoupling techniques in controlling a robot&rsquo;s movement and ability to interact with its surrounding. As they explained, hierarchy-based decoupled and semi-coupled control methods have been ineffective due to a &ldquo;lack of coordination between the arm and legs, error propagation across modules, and slow, non-smooth and unnatural motions.&rdquo;</p><p>The unified policy was based on a machine-learning training concept called <strong>reinforcement learning (RL)</strong>. This method simply rewards desirable behaviors and punishes undesirable ones. Through RL, the robot would learn by trial and error to perceive and analyze its environment and act accordingly.</p><p dir="ltr">The study comprised simulation experiments, which were then transferred into real-world experimentation. However, three distinct issues made it challenging beyond the standard RL model of training in simulation and transferring to the real world:</p><ol><li dir="ltr"><p dir="ltr">Multiple degrees of freedom (12 DoFs in the quadruped robot and 6 DoF in the attached arm)</p></li><li dir="ltr"><p dir="ltr">Conflicting objectives in mobility and weight balance</p></li><li dir="ltr"><p dir="ltr">Dependency between manipulation and locomotion&nbsp;&nbsp;</p></li></ol><p dir="ltr">This is why Fu et al. proposed adding&nbsp;regularized online adaptation&nbsp;to bridge the so-called Sim2Real gap - i.e., the information gap between the skills acquired in simulation and those transferred to the real-world system; also referred to as the&nbsp;realizability gap. Regularization is a machine learning technique that helps calibrate the model and minimize the deviation between the model and the dataset, thus minimizing the realizability gap. This adaptation helps eliminate the conventional, two-phase, teacher-student scheme used in previous Sim2Real transfer applications. This is where a &ldquo;teacher network&rdquo; is trained by RL in simulation using full-state information, and a &ldquo;student network&rdquo; attempts to imitate it in real life based on partial information from onboard observations.&nbsp;<iframe id="6389f2641a0dc20008fdb9d3" width="250" class="specs-iframe" height="400" src="https://wevolver.com/iframe/specs/interbotix-widowx-250-6dof?iframe=true" frameborder="0" scrolling="true" data-lf-form-tracking-inspected-kn9eq4roawkarlvp="true" data-lf-yt-playback-inspected-kn9eq4roawkarlvp="true" data-lf-vimeo-playback-inspected-kn9eq4roawkarlvp="true"></iframe></p><p dir="ltr">The researchers also tackled the dependency issue by proposing the concept of&nbsp;advantage mixing. Advantage functions describe how much a particular action is a good or bad decision given a specific state. Knowing that locomotion tasks mostly rely on leg activity and manipulation tasks on arm activity, mixing the advantage functions of both locomotion and manipulation can help encourage and boost the robot&rsquo;s learning of the unified policy.</p><h2 dir="ltr">A Low-Cost Legged-Robot with Whole-Body Control</h2><p dir="ltr">The researchers presented a simple design for the legged manipulator made of the following elements:</p><ol><li dir="ltr"><p dir="ltr">A 4-legged robot platform</p></li><li dir="ltr"><p dir="ltr">A robot arm on top of the platform</p></li><li dir="ltr"><p dir="ltr">An RGB camera positioned next to the gripper of the arm</p></li><li dir="ltr"><p dir="ltr">An onboard battery to power the arm and the quadruped robot</p></li></ol><h2 dir="ltr">Robust Unified Policy in Simulation and Real World</h2><p dir="ltr">With this design, they were able to show in the simulation experiments an increase in whole-body coordination as the legs helped the arm reach beyond its own workspace by bending to reach lower objects (e.g. picking up a cup) and standing up high to reach higher objects (e.g. wiping a whiteboard). Similarly, the arm helped the legs maintain balance even under relatively large disturbances (e.g. running with an initial speed of 1 m/s). This demonstrates a solid unified policy with a high survival rate.</p><p dir="ltr">Advantage mixing also helped the policy focus on each task and then combine them to induce a mechanism that boosts training. Furthermore, the proposed regularized adaptation resulted in improved performance and better prediction of the environment, as errors at the level of the gripper dropped by 20% compared to other systems, such as the two-phase teacher-student scheme.</p><p dir="ltr">As for the real-world experiments, the researchers studied three activities: teleoperation using joysticks, closed-loop control through RGB vision-guided tracking, and open-loop response to human demonstrations.</p><ul><li dir="ltr"><p dir="ltr">For teleoperation, the researchers used two joysticks to command the end-effector (gripper) to reach particular points even beyond the space of training. They found out that both the leg and arm joints coordinated to help the end-effector reach those points.</p></li><li dir="ltr"><p dir="ltr">During vision-guided tracking, the RGB camera was used alongside the joystick control to provide visual feedback to the robot and improve its picking tasks. It showed repeated success on both easy and hard tasks.</p></li><li dir="ltr"><p dir="ltr">In the open-loop control experiments, the researchers tested the coupling of the robot&rsquo;s agile locomotion with its dynamic arm movement. They commanded the end-effector to follow a predefined path while the robot was walking. The results showed dynamic coordination even while walking on an uneven grass area.</p></li></ul><p dir="ltr"><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/zhpI3JTqQGI?&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><br></p><h2 dir="ltr">How the Unified Policy Approach be Further Expanded</h2><p dir="ltr">The success of the unified policy opens up a new scope of coordinated robotic movement. As the researchers suggested, those results are merely preliminary. Further exciting research is expected to stem from this work, such as enabling the robot to climb on a table using its front legs to pick something up from atop the table. Another suggestion was to mount a camera onto the center of the torso, providing different visual feedback, which would pave the road for vision-based policy learning.</p><p dir="ltr"><a href="https://manipulation-locomotion.github.io/" target="_blank">Read the full study and learn more about the exciting research of Fu et al. by visiting their dedicated website.</a></p><p dir="ltr">1.&nbsp;<span style="background-color: transparent;">Zipeng , F., Cheng, X. and Pathak, D. (2022) &ldquo;Deep Whole-Body Control: Learning a Unified Policy for Manipulation and Locomotion ,&rdquo;&nbsp;</span><em style="background-color: transparent;">CoRL 2022 Conference Paper</em><span style="background-color: transparent;">&nbsp;[Preprint].&nbsp;</span></p><p><br></p><p><br></p>

Could this be the time we say goodbye to robotic awkwardness and hello to seamless motion? This article highlights the groundbreaking concept of whole-body control, a coordinated robotic motion technique that resembles natural movement. Imagine a robot smoothly navigating its environment, performing tasks with grace and precision. This innovative approach allows for improved efficiency and greater versatility in real-world settings.


Unifying Robotic Arm and Leg Movement for Whole-Body Control

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


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

<p id="isPasted">The results build on evidence that providing richer information during artificial intelligence (AI) training can make autonomous robots more adaptive to new situations, improving their safety and effectiveness.</p><p>Adding descriptions of a tool&rsquo;s form and function to the training process for the robot improved the robot&rsquo;s ability to manipulate newly encountered tools that were not in the original training set. A team of mechanical engineers and computer scientists presented the new method, <a href="https://openreview.net/pdf?id=nPw7jaGBrCG" target="_blank">Accelerated Learning of Tool Manipulation with LAnguage</a>, or ATLA, at the Conference on Robot Learning on Dec. 14.</p><p><span class="fr-video fr-deletable fr-fvc fr-dvb fr-draggable" contenteditable="false" draggable="true"><iframe src="https://player.vimeo.com/video/781953242" 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" style="width: 764px; height: 492px;"></iframe></span></p><p id="isPasted">Robotic arms have great potential to help with repetitive or challenging tasks, but training robots to manipulate tools effectively is difficult: Tools have a wide variety of shapes, and a robot&rsquo;s dexterity and vision are no match for a human&rsquo;s.</p><p>&ldquo;Extra information in the form of language can help a robot learn to use the tools more quickly,&rdquo; said study coauthor <a href="https://engineering.princeton.edu/faculty/anirudha-majumdar" target="_blank">Anirudha Majumdar</a>, an assistant professor of <a href="https://mae.princeton.edu/" target="_blank">mechanical and aerospace engineering</a> at Princeton who leads the <a href="https://irom-lab.princeton.edu/majumdar" target="_blank">Intelligent Robot Motion Lab</a>.</p><p>The team obtained tool descriptions by querying GPT-3, a large language model released by OpenAI in 2020 that uses a form of AI called deep learning to generate text in response to a prompt. After experimenting with various prompts, they settled on using &ldquo;Describe the [feature] of [tool] in a detailed and scientific response,&rdquo; where the feature was the shape or purpose of the tool.</p><p>&ldquo;Because these language models have been trained on the internet, in some sense you can think of this as a different way of retrieving that information,&rdquo; more efficiently and comprehensively than using crowdsourcing or scraping specific websites for tool descriptions, said <a href="https://engineering.princeton.edu/faculty/karthik-narasimhan" target="_blank">Karthik Narasimhan</a>, an assistant professor of <a href="https://www.cs.princeton.edu/people/profile/karthikn" target="_blank">computer science</a> and coauthor of the study. Narasimhan is a lead faculty member in Princeton&rsquo;s <a href="https://princeton-nlp.github.io/" target="_blank">natural language processing (NLP) group</a>, and contributed to the original GPT language model as a visiting research scientist at OpenAI.</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%2F1671785093867-pushing_tasks.gif" style="width: 300px;" class="fr-fic fr-dib"><span class="fr-inner">Princeton researchers have found that human-language descriptions of tools can accelerate the learning of a simulated robotic arm lifting and using a variety of tools. Animation by the researchers; GIF by Neil Adelantar&nbsp;</span></span></span></p><p>This work is the first collaboration between Narasimhan&rsquo;s and Majumdar&rsquo;s research groups. Majumdar focuses on developing AI-based policies to help robots &mdash; <a href="https://engineering.princeton.edu/news/2020/11/17/machine-learning-guarantees-robots-performance-unknown-territory" target="_blank">including flying and walking robots</a> &mdash; generalize their functions to new settings, and he was curious about the potential of recent &ldquo;massive progress in natural language processing&rdquo; to benefit robot learning, he said.</p><p>For their simulated robot learning experiments, the team selected a training set of 27 tools, ranging from an axe to a squeegee. They gave the robotic arm four different tasks: push the tool, lift the tool, use it to sweep a cylinder along a table, or hammer a peg into a hole. The researchers developed a suite of policies using machine learning training approaches with and without language information, and then compared the policies&rsquo; performance on a separate test set of nine tools with paired descriptions.</p><p>This approach is known as meta-learning, since the robot improves its ability to learn with each successive task. It&rsquo;s not only learning to use each tool, but also &ldquo;trying to learn to understand the descriptions of each of these hundred different tools, so when it sees the 101st tool it&rsquo;s faster in learning to use the new tool,&rdquo; said Narasimhan. &ldquo;We&rsquo;re doing two things: We&rsquo;re teaching the robot how to use the tools, but we&rsquo;re also teaching it English.&rdquo;</p><p>The researchers measured the success of the robot in pushing, lifting, sweeping and hammering with the nine test tools, comparing the results achieved with the policies that used language in the machine learning process to those that did not use language information. In most cases, the language information offered significant advantages for the robot&rsquo;s ability to use new tools.</p><p>One task that showed notable differences between the policies was using a crowbar to sweep a cylinder, or bottle, along a table, said <a href="https://zhiyir.mycpanel.princeton.edu/" target="_blank">Allen Z. Ren</a>, a Ph.D. student in Majumdar&rsquo;s group and lead author of the research paper.</p><p>&ldquo;With the language training, it learns to grasp at the long end of the crowbar and use the curved surface to better constrain the movement of the bottle,&rdquo; said Ren. &ldquo;Without the language, it grasped the crowbar close to the curved surface and it was harder to control.&rdquo;</p><p>The research was supported in part by the Toyota Research Institute (TRI), and is part of a larger TRI-funded project in Majumdar&rsquo;s research group aimed at improving robots&rsquo; ability to function in novel situations that differ from their training environments.</p><p>&ldquo;The broad goal is to get robotic systems &mdash; specifically, ones that are trained using machine learning &mdash; to generalize to new environments,&rdquo; said Majumdar. Other TRI-supported work by his group has addressed <a href="https://arxiv.org/abs/2202.05894" target="_blank">failure prediction</a> for vision-based robot control, and used an <a href="https://arxiv.org/abs/2107.06353" target="_blank">&ldquo;adversarial environment generation&rdquo;</a> approach to help robot policies function better in conditions outside their initial training.</p><p>The article, <em>Leveraging language for accelerated learning of tool manipulation</em>, was presented Dec. 14 at the Conference on Robot Learning. Besides Majumdar, Narasimhan and Ren, coauthors include Bharat Govil, Princeton Class of 2022, and Tsung-Yen Yang, who completed a Ph.D. in electrical engineering at Princeton this year and is now a machine learning scientist at Meta Platforms Inc.</p><p>In addition to TRI, support for the research was provided by the U.S. National Science Foundation, the Office of Naval Research, and the School of Engineering and Applied Science at Princeton University through the generosity of William Addy &rsquo;82.</p>

Princeton researchers have found that human-language descriptions of tools can accelerate the learning of a simulated robotic arm lifting and using a variety of tools. Animation by the researchers; GIF by Neil Adelantar

Exploring a new way to teach robots, Princeton researchers have found that human-language descriptions of tools can accelerate the learning of a simulated robotic arm lifting and using a variety of tools.