Image of unexploded ordnance (UXO) on sea floor.

Learn all about underwater mines and explosive ordnance disposal, the risks involved, and how ROVs can help while providing extra safety and speed for divers.

Everything You Need To Know About Underwater Mines

Explore the distinct functionalities, advantages, and industries driving the innovations of AC and DC motors. Delve into the technologies that shape our modern world, focusing on recent advancements and real-world applications. Understand their role in future engineering challenges and opportunities.

AC vs DC Motor: Unpacking Their Engineering Applications and Innovations

In the dynamic landscape of automotive manufacturing, innovation has focused mostly on delivering greater efficiency, automation, sustainability, and flexibility. One such innovation that is contributing to this transformation is Large Format Additive Manufacturing (LFAM), a cutting-edge technology that promises to reshape how vehicles are prototyped, manufactured, and even customized. Let’s delve into the advantages and primary applications of LFAM within the automotive industry.MAIN APPLICATIONS OF LFAM IN THE AUTOMOTIVE INDUSTRYPrototyping and Tooling: LFAM expedites the prototyping phase, allowing engineers to quickly test and refine automotive designs before committing to costly tooling. Additionally, LFAM produces durable, customized composite tooling (e.g., jigs, fixtures, autoclave cure tools, molds) that enhance manufacturing efficiency.Customized Components: Automotive manufacturers are increasingly leveraging LFAM to produce bespoke and low-volume parts tailored to customer preferences. This personalization extends beyond aesthetics to functional components, such as lightweight brackets and interior trim.Structural Elements: LFAM is employed in manufacturing large structural components like vehicle chassis and body panels. By leveraging lightweight materials and optimized designs, automakers can achieve superior strength-to-weight ratios, enhancing vehicle performance and fuel efficiency.Spare Parts Production: LFAM enables on-demand production of legacy or discontinued automotive parts. This capability reduces inventory costs associated with maintaining a vast stockpile of spare parts and ensures timely availability for customers.Innovative Vehicle Concepts: LFAM opens avenues for revolutionary vehicle designs, including electric and autonomous vehicles. Its ability to fabricate complex shapes and integrate functional features directly into components facilitates the realization of next-generation automotive concepts.Large Format Additive Manufacturing represents a paradigm shift in automotive manufacturing, offering unmatched design freedom, production agility, and material versatility. As automakers strive for sustainable practices and accelerated innovation cycles, LFAM emerges as a transformative technology poised to shape the future of mobility.By harnessing the full potential of LFAM, automotive manufacturers can expedite time-to-market, reduce costs, and unlock unprecedented opportunities for vehicle design and performance. Caracol’s robotic LFAM platforms Heron AM are used to realize automotive composites applications including autoclave tools, functional and aesthetic mock-ups, car’s prototypes, and customized finished parts.

In the dynamic landscape of automotive manufacturing, innovation has focused mostly on delivering greater efficiency, automation, sustainability, and flexibility. Large Format 3D printing is a cutting-edge technology that promises to reshape how vehicles are prototyped, manufactured and customized.

THE TRANSFORMATIVE POWER OF LARGE FORMAT 3D PRINTING IN AUTOMOTIVE

Revolutionizing Industrial Manufacturing with Robotic Large Format Additive Manufacturing (LFAM)In the realm of industrial technology, 3D printing has made remarkable strides, particularly in large-scale applications across diverse sectors like aerospace, automotive, energy, and construction. A key breakthrough in this arena is Robotic Large Format Additive Manufacturing (LFAM), a game-changing technology that merges 3D printing with robotic systems to create massive and intricate objects.Understanding Robotic LFAMRobotic LFAM integrates 3D printing processes with robotic movement support, enabling the production of objects at unprecedented scales beyond traditional 3-axis frameworks.&nbsp;The synergy between additive manufacturing and robotics unlocks new possibilities. Industrial robots, with their varied sizes and arm extensions, facilitate the production of large, intricate parts with flexibility. Caracol's Heron AM platform allows for the integration of different robotic arms with varied printing heads, to suit specific customer needs. This cutting-edge approach allows for the creation of complex, large-scale parts—ranging from meters in size—essential for industries where part dimensions and production times are critical.The Advantages of FGF 3D Printing&nbsp;Robotic LFAM systems like Caracol's Heron AM platform utilize Fused Granulate Fabrication (FGF), an extrusion-based technology using thermoplastic pellets. FGF leverages cost-effective thermoplastics sourced from various materials, including composites and recycled plastics, emphasizing a circular economy approach. This method enables customization of material recipes for optimized 3D printing processes that ensure quality and efficiency in the LFAM process.Flexibility and Simplified ProgrammingRobotic LFAM offers unparalleled flexibility in design and production. The anthropomorphic movements of robots enable the realization of daring shapes and geometries, expanding the potential of 3D printing technologies. Heron AM printing area and system configuration can be set according to the final applications’ dimensions; while Caracol's software solutions Eidos Manufacturing simplify robotics programming, ensuring seamless integration with the 3D printing process for efficient production.Read more on how robotic LFAM is transforming the landscape of industrial manufacturing, offering scalability, efficiency, and sustainability at <a href="https://www.caracol-am.com/resources/knowledge-hub/what-is-robotic-lfam" target="_blank" rel="noopener noreferrer">this link</a> and connect with Caracol’s experts to explore the potential of large-scale additive manufacturing for your business.

Learn what Large Format Additive Manufacturing is and understand key aspects such as robotics and FGF-based 3D printing technique. In this video, you will understand why Caracol’s LFAM platforms combine 6+ axes robotics with 3D printing, the reasons, and actions behind the development of an all-in-one solution integrating hardware and software to enhance efficiency, and the great advantages in terms of flexibility and sustainability of a modular solution such as Heron AM, to configure according to the specific application needs. Starting from pellets-form thermoplastic materials, LFAM can deliver tangible benefits to your industrial large-scale parts production.

3D printing from pellets meets robotics to deliver tangible benefits on parts and applications across sectors. With no limits in scale, shape or material, Caracol's LFAM technology enhances efficiency through continuity, flexiblity, and sustainability to your production.

LFAM, FGF, and robotics... what exactly are we talking about?

The <a href="https://wevlv.co/mik-2024" target="_blank" rel="noopener noreferrer">Vision, Robotics &amp; Motion Trade</a> event is an annual event held in Brabanthallen in 's-Hertogenbosch, the Netherlands, that offers a complete overview of solutions for smart automation and digitalisation opportunities. The next two-day event will be held between the 5th and 6th of June, 2024The event aims to provide insights into robotics, vision, and motion implementation in the manufacturing sector, enabling engineers and technical specialists within the MKB/OEM high-tech and manufacturing industry to learn from industry leaders about how to address challenges, such as consumer demands and increasing automation. &nbsp;With over 130 exhibitors, including manufacturers, system integrators, suppliers, and advisors in attendance, this event promises to be more exciting than the previous year.<img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNzE1NzYwNzU1Mzk4LTUyOTYxNjE2OTE2XzhjNmE4NTA4ZDZfay5qcGciLCJlZGl0cyI6eyJyZXNpemUiOnsid2lkdGgiOjk1MCwiZml0IjoiY292ZXIifX19" style="width: 100%px;" class="fr-fic fr-dib">This year's event will continue to explore automation in the manufacturing industry. This year's trade show will be expanded to include a new pillar: Digital Innovation. This pillar focuses on digitization opportunities and solutions for companies in the manufacturing industry. “We see that digitalization and automation always go hand in hand, without digitalization you can hardly automate. Therefore, it can hardly exist separately. That makes the combination at the fair very interesting,” says Timo van Leent, manager of the Events Department at Mikrocentrum. “Entrepreneurs can get informed about the latest trends. Moreover, the stories of others are often inspiring.”Exhibitors will demonstrate the latest trends and technology in smart production automation with use cases, and visitors will be able to engage directly with manufacturing companies and hear about cutting-edge technological developments and applications.<h2 dir="ltr">Robotics Demonstrations and AI Applications</h2>One of the most engaging demo areas of the floor plan hosts a variety of robotic systems in full operation. &nbsp;Industrial robots and collaborative robots will be exhibited, performing specific tasks and working with human operators. Experts will showcase the versatility and efficiency of these robotic systems in performing complex operations and executing tasks with speed and precision.Furthermore, the demo areas will have several AI applications that utilize machine learning and computer vision technologies. Attendees can witness intelligent algorithms analyzing data and AI-powered vision systems identifying defects in real time. These demos pinpoint the power of AI in manufacturing and other industries.Industry experts in robotics and AI will assist with suitable solutions and implementations on the exhibition floor. The various speakers will guide with smart and practical digitalization solutions and discuss the potential impact of advancements in AI and robotics on manufacturing industries. The demo areas will have interactive workshops to allow attendees to engage in discussions with authority figures in the field.&nbsp;<h2 dir="ltr">The Significance of this Event to the Manufacturing Industry</h2>The importance of this event lies in its approach to addressing manufacturers' needs through new technological innovations. Manufacturers' problems are multifaceted, from rising consumer demands to product delivery. In today's fast-paced world, where efficiency, quality control, and sustainability are the core of production, manufacturers face an uphill task in adapting and meeting demands.Charlotte de Jong, program manager of Vision, Robotics, and Motion at Mikrocentrum, said more and more manufacturing companies want to automate for various reasons. The shortage in the labor market exemplifies one of these reasons. The shortage means companies have to find ways to work more efficiently. Robotics and AI offer innovative solutions to address this global challenge. Robotics and AI can help in automating repetitive tasks with high precision and speed, while they help to streamline manufacturing processes and minimize waste, thereby maximizing output and reducing operational costs.<img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNzE1NzYwNzcyMDQ3LTUyOTYxMDE2MDc3X2YwZjMxMTkyN2Rfay5qcGciLCJlZGl0cyI6eyJyZXNpemUiOnsid2lkdGgiOjk1MCwiZml0IjoiY292ZXIifX19" style="width: 100%px;" class="fr-fic fr-dib">AI also ensures that manufacturers maintain a high level of quality and meet regulatory standards. Leveraging advanced solutions, such as automated inspection systems, can help to identify defects accurately, ensuring only products of the highest quality are manufactured and delivered to consumers. When the risk of defects is reduced, manufacturers ensure quality control and uphold their reputation.Furthermore, robotics and AI can enable greater flexibility. In today's consumer market, manufacturers must adapt quickly to consumer dynamism and market demands. Traditional manufacturing often slacks in responding to rapid changes, leading to inefficiencies and missed opportunities. Robotics and AI, in conjunction with human efforts, will facilitate quick reconfiguration of production lines and offer analytical insights that will help manufacturers adjust to demand fluctuations and optimize production schedules for greater results.Finally, this power-packed event will make clear how these transformative solutions can bring sustainability throughout the production. With more eyes now on environmental impact, manufacturers are under immense pressure to abide by sustainable practices during production. Robotics and AI can help manufacturers achieve sustainability goals by reducing energy consumption and minimizing waste. These advanced solutions can regulate energy usage and repurpose materials to reduce waste and costs.&nbsp;AI will unlock new levels of innovation and productivity, and the manufacturing industry can greatly benefit from this transformative solution. The upcoming Vision, Robotics, and Motion event hopes to create more awareness of the advantages of robotics and AI and how their implementation in the manufacturing sector will increase competitiveness, enhance efficiency, and enable flexibility.&nbsp;<h2 dir="ltr">Digital Innovation: A New Pillar of this Year’s Event</h2>A new addition to this year’s event is digital innovation, which will be one of the key highlight of the trade fair. This new pillar will reflect the increasing importance of digitalization in manufacturing. It focuses on digitalization opportunities and solutions for manufacturing companies.&nbsp;Manager of the Events Department, Timo Van Leent, said digitalization and automation are intertwined. "Without digitalization, you can't automate," he said. That's what makes the upcoming event in the Netherlands interesting. Attendees can learn about the latest trends and discover innovative technologies that will revolutionize manufacturing. From maintenance to logistics, digital innovation will unlock new opportunities for enhanced productivity, enabling more agile and responsive manufacturing lines.Leent reveals that while there's a clamor to adopt digitalization and automation in manufacturing, small manufacturing companies are slightly disadvantaged because they need higher budgets to experiment. However, he states that these companies don't need to follow the latest innovation trends. They can gain insights at the Vision, Robotics, and Motion trade fair into technologies that can help them achieve digitalization and automation at their level.Attending engineers will gain insight into digitalization and automation from exhibitors and have the opportunity to build a network. De Jong states that networking is fundamental at the event. "Visitors get to know many people while having the chance to check out some technologies." In addition to the exhibition and the demos, Manufacturers and engineers will be given an extensive lecture on the use of the software. Learning and discovery are essential aspects of the "Digital Innovation" pillar.<h2 dir="ltr">Embracing Digital Innovation in Other Industries</h2>Beyond manufacturing, the technologies showcased at this year’s event will also influence other industries, such as healthcare and agriculture. In healthcare, AI-powered robotic-assisted surgeries are becoming prevalent in patient care. In agriculture, robotics and AI are enhancing sustainable crop yields.&nbsp;The latest robotics innovations are poised to transform these domains globally, focusing on efficiency and productivity. Attendees can learn of their practical applications and hope for tangible benefits beyond manufacturing.&nbsp;<h2 dir="ltr">Vital Learning Opportunities</h2>Trade fairs like this will serve as a crucial learning platform for students, professionals, entrepreneurs, and even robotics enthusiasts. Attendees will have hands-on learning, equipped with the skills and knowledge required to thrive in this fast-paced, technologically driven world.&nbsp;<h2 dir="ltr" id="isPasted">Preparing for the Future</h2>The innovations showcased at this year’s Vision, Robotics, and Motion event offer a glimpse into the exciting possibilities of automation in manufacturing and beyond. This event will prepare us for the tomorrow we anticipate—a prosperous and sustainable future that will harness the full potential of robotics and AI.&nbsp;What:&nbsp;Vision, Robotics &amp; Motion TradeWhen:&nbsp;5th and 6th of June, 2024Where:&nbsp;'s-Hertogenbosch, the Netherlands,&nbsp;<a href="https://wevlv.co/mik-2024" target="_blank" rel="noopener noreferrer" class="button-main-action">Register to attend for free here!</a>

June 5th-6th, 2024, 's-Hertogenbosch, Netherlands

Vision, Robotics, & Motion Event 2024: The Future of Smart Production, Automation and Digitalization

In this episode, we discuss how a team from Carnegie Mellon University spearheading non-invasive brain computer interface solutions has had a significant breakthrough in improving their accuracy.

Podcast: All The Brain Chip Implant Benefits & None of The Risks

Parents spend roughly 300 hours each year cleaning up after their kids. That’s nearly two weeks that could be otherwise spent doing, well, anything else. Researchers in Carnegie Mellon University’s <a href="https://www.meche.engineering.cmu.edu/index.html" rel="noopener" target="_blank" id="isPasted">Department of Mechanical Engineering</a>, in collaboration with Google DeepMind and the University of Washington, are hopeful that parents will regain tidying time and then some with <a href="https://linchangyi1.github.io/LocoMan/" rel="noopener" target="_blank">LocoMan</a>, a four-legged version of the Jetsons’ beloved maid “Rosey the Robot,” that could lend a hand in the not-so-distant future.<iframe width="640" height="360" src="https://www.youtube.com/embed/d-GLiRPiHIE?&amp;wmode=opaque&amp;enablejsapi=1&amp;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">&nbsp;&nbsp;&nbsp;</iframe>Quadrupedal robots, designed to move on four legs much like a dog, are already found leading search and rescue missions, surveying construction sites, and moving parts around factories. Their ability to navigate complex environments with a low center of gravity and light weight make quadrupedal robots a top contender for household use, too. Until now, the missing piece was the dexterity and versatility skills necessary for them to be efficient in daily life.By installing two custom-designed, lightweight manipulators (think robot arms) onto the front legs of a quadrupedal robot, <a href="https://engineering.cmu.edu/directory/bios/zhao-ding.html" target="_blank">Ding Zhao</a> has created a robot versatile enough to open doors, pour drinks, plug a phone in to charge, and—perhaps best of all—clean up after your on-the-go toddlers.<blockquote>While many researchers work on humanoid robots that look like humans, with potential to directly learn from humans or even replace humans in some tasks, we focus on robots that are complementary to humans to team up with them.<cite>Ding Zhao, Associate Professor, Mechanical Engineering</cite></blockquote>“While many researcher work on humanoid robots that look like humans, with potential to directly learn from humans or even replace humans in some tasks, we focus on robots that are complementary to humans to team up with them,” said Zhao, associate professor of mechanical engineering.Unlike humanoid robots that come at a high cost and hefty weight, Zhao’s manipulators can be added to existing lower-cost quadrupedal robots and are compact, lightweight, and easy to fabricate, making them an accessible technology.<img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNzE1NzM1MTQ3ODg3LTA1MDktZW0taG9tZS1oZWxwZXItcm9ib3QucG5nIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" style="width: 100%px;" class="fr-fic fr-dib">The core idea is to leverage the leg as an integral component of the “robotic arm” for executing 6D pose manipulation.“We used four commercially available servos and 3D printed a few other parts to bring them to life,” Changyi Lin, a Ph.D. candidate in the CMU Safe AI Lab and the first author of the paper explained.Zhao’s team introduced the robot to a class of five preschool students at The Cyert Center for Early Education. After explaining to children how robots work and what we use them for, the children had the opportunity to interact with Zhao’s robot. The students were impressed with the robot’s capabilities and didn’t want it to leave their classroom.<img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNzE1NzM1MTc3MDc1LTA1MTMtZW0yLWhvbWUtaGVscGVyLXJvYm90LnBuZyIsImVkaXRzIjp7InJlc2l6ZSI6eyJ3aWR0aCI6OTUwLCJmaXQiOiJjb3ZlciJ9fX0=" style="width: 100%px;" class="fr-fic fr-dib">Ding Zhao (center), with children from the Cyert Center, interacting with the home-helper robot.The group continues to develop LocoMan to perform tasks autonomously by integrating advanced perception and intelligence planning capabilities into its existing whole-body controller.“We anticipate that robots like LocoMan will be ready to help out at home, or in daycares, within the next three years,” said Zhao. “It is a pleasant job just to imagine that I may no longer need to handle my kids’ messy fun.”This work will be presented at the loco-manipulation workshop and the manipulation skills workshop of the 2024 IEEE International Conference on Robotics and Automation.

New locomotion and manipulation technology from Ding Zhao’s lab will enable four-legged robots to lend a hand in the not-so-distant future.

Manipulation technology makes home-helper robot possible

Introducing SKYRON 2, the latest innovation in flying robotic inspection explicitly tailored for the non-destructive testing industry. Building upon SKYRON's visual and contact-based inspection tools, SKYRON 2 offers advanced capabilities for comprehensive assessments. Safety during flight operations is prioritized with SKYRON 2. A full product release on SKYRON 2 will happen later this May.

The latest innovation in inspection flying robots explicitly tailored for the non-destructive testing (NDT) industry.

SKYRON 2 Release

Imagine a slime-like robot that can seamlessly change its shape to squeeze through narrow spaces, which could be deployed inside the human body to remove an unwanted item.While such a robot does not yet exist outside a laboratory, researchers are working to develop reconfigurable soft robots for applications in health care, wearable devices, and industrial systems.But how can one control a squishy robot that doesn’t have joints, limbs, or fingers that can be manipulated, and instead can drastically alter its entire shape at will? MIT researchers are working to answer that question.They developed a control algorithm that can autonomously learn how to move, stretch, and shape a reconfigurable robot to complete a specific task, even when that task requires the robot to change its morphology multiple times. The team also built a simulator to test control algorithms for deformable soft robots on a series of challenging, shape-changing tasks.<iframe width="640" height="360" src="https://www.youtube.com/embed/IieJM8wYUKo?&amp;wmode=opaque&amp;rel=0&amp;enablejsapi=1&amp;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>Their method completed each of the eight tasks they evaluated while outperforming other algorithms. The technique worked especially well on multifaceted tasks. For instance, in one test, the robot had to reduce its height while growing two tiny legs to squeeze through a narrow pipe, and then un-grow those legs and extend its torso to open the pipe’s lid.While reconfigurable soft robots are still in their infancy, such a technique could someday enable general-purpose robots that can adapt their shapes to accomplish diverse tasks.“When people think about soft robots, they tend to think about robots that are elastic, but return to their original shape. Our robot is like slime and can actually change its morphology. It is very striking that our method worked so well because we are dealing with something very new,” says Boyuan Chen, an electrical engineering and computer science (EECS) graduate student and co-author of a <a href="https://arxiv.org/pdf/2401.13231" target="_blank">paper on this approach</a>.Chen’s co-authors include lead author Suning Huang, an undergraduate student at Tsinghua University in China who completed this work while a visiting student at MIT; Huazhe Xu, an assistant professor at Tsinghua University; and senior author Vincent Sitzmann, an assistant professor of EECS at MIT who leads the Scene Representation Group in the Computer Science and Artificial Intelligence Laboratory. The research will be presented at the International Conference on Learning Representations.<h2>Controlling dynamic motion</h2>Scientists often teach robots to complete tasks using a machine-learning approach known as reinforcement learning, which is a trial-and-error process in which the robot is rewarded for actions that move it closer to a goal.This can be effective when the robot’s moving parts are consistent and well-defined, like a gripper with three fingers. With a robotic gripper, a reinforcement learning algorithm might move one finger slightly, learning by trial and error whether that motion earns it a reward. Then it would move on to the next finger, and so on.But shape-shifting robots, which are controlled by magnetic fields, can dynamically squish, bend, or elongate their entire bodies.<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1715580226145-Ditto-gym-1.gif" style="width: 100%px;" class="fr-fic fr-dib">The researchers built a simulator to test control algorithms for deformable soft robots on a series of challenging, shape-changing tasks. Here, a reconfigurable robot learns to elongate and curve its soft body to weave around obstacles and reach a target. Image: Courtesy of the researchers“Such a robot could have thousands of small pieces of muscle to control, so it is very hard to learn in a traditional way,” says Chen.To solve this problem, he and his collaborators had to think about it differently. Rather than moving each tiny muscle individually, their reinforcement learning algorithm begins by learning to control groups of adjacent muscles that work together.Then, after the algorithm has explored the space of possible actions by focusing on groups of muscles, it drills down into finer detail to optimize the policy, or action plan, it has learned. In this way, the control algorithm follows a coarse-to-fine methodology.“Coarse-to-fine means that when you take a random action, that random action is likely to make a difference. The change in the outcome is likely very significant because you coarsely control several muscles at the same time,” Sitzmann says.To enable this, the researchers treat a robot’s action space, or how it can move in a certain area, like an image.Their machine-learning model uses images of the robot’s environment to generate a 2D action space, which includes the robot and the area around it. They simulate robot motion using what is known as the material-point-method, where the action space is covered by points, like image pixels, and overlayed with a grid.The same way nearby pixels in an image are related (like the pixels that form a tree in a photo), they built their algorithm to understand that nearby action points have stronger correlations. Points around the robot’s “shoulder” will move similarly when it changes shape, while points on the robot’s “leg” will also move similarly, but in a different way than those on the “shoulder.”In addition, the researchers use the same machine-learning model to look at the environment and predict the actions the robot should take, which makes it more efficient.<h2>Building a simulator</h2>After developing this approach, the researchers needed a way to test it, so they created a simulation environment called DittoGym.DittoGym features eight tasks that evaluate a reconfigurable robot’s ability to dynamically change shape. In one, the robot must elongate and curve its body so it can weave around obstacles to reach a target point. In another, it must change its shape to mimic letters of the alphabet.<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1715580352121-ezgif-1-783e5d8fbe.gif" style="width: 100%px;" class="fr-fic fr-dib">In this simulation, the reconfigurable soft robot, trained using the researchers' control algorithm, must change its shape to mimic objects, like stars, and the letters M-I-T. Image: Courtesy of the researchers“Our task selection in DittoGym follows both generic reinforcement learning benchmark design principles and the specific needs of reconfigurable robots. Each task is designed to represent certain properties that we deem important, such as the capability to navigate through long-horizon explorations, the ability to analyze the environment, and interact with external objects,” Huang says. “We believe they together can give users a comprehensive understanding of the flexibility of reconfigurable robots and the effectiveness of our reinforcement learning scheme.”Their algorithm outperformed baseline methods and was the only technique suitable for completing multistage tasks that required several shape changes.“We have a stronger correlation between action points that are closer to each other, and I think that is key to making this work so well,” says Chen.While it may be many years before shape-shifting robots are deployed in the real world, Chen and his collaborators hope their work inspires other scientists not only to study reconfigurable soft robots but also to think about leveraging 2D action spaces for other complex control problems.

A new machine-learning technique can train and control a reconfigurable soft robot that can dynamically change its shape to complete a task. The researchers, from MIT and elsewhere, also built a simulator that can evaluate control algorithms for shape-shifting soft robots. Image: Courtesy of the researchers; MIT News

A new algorithm learns to squish, bend, or stretch a robot’s entire body to accomplish diverse tasks like avoiding obstacles or retrieving items.

A better way to control shape-shifting soft robots

In today’s manufacturing sectors, companies are continually confronted with various pressing challenges. Labor shortages and rising labor costs require innovative solutions to maintain productivity with less staff.In addition, relentless inflation continues to exert pressure on raw material costs, squeezing margins. Manufacturers are also racing against tight production timelines, driven by the imperative of shorter time-to-market windows—a direct consequence of escalating global competition.To overcome these pervasive issues, many executives and factory managers are turning to&nbsp;<a href="https://www.creaform3d.com/blog/learn-how-to-leverage-the-potential-of-automation/" target="_blank">automation and metrology solutions</a>. However, they want to leverage these technologies further to maximize their investments. So, what are the next possible steps?This article provides insights into how manufacturing companies can harness the potential of 3D scanners within automated environments. It also offers guidance for adopting approaches that balance initial technology investments with future scalability toward end-to-end automation.<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1715330446582-MetraSCAN-R-BLACK-HD_car-frame-inspection_047-2048x1365.webp" style="width: 100%px;" class="fr-fic fr-dib"><h2 id="isPasted">Manufacturers: Still Hesitant about Automation</h2>Despite the widespread recognition of automation as an indispensable factor in modern manufacturing, manufacturers remain hesitant to fully embrace automated production lines. This reluctance is rooted in organizational culture and operational considerations.<h3>The Perception of Automation as Overly Complex</h3>For many manufacturers, the perceived complexity of automation serves as a deterrent. The notion that automation requires a radical overhaul of existing processes and equipment fosters a fear of the unknown. There’s a prevalent concern that introducing sophisticated systems like robots and cobots will lead to job redundancies and a steep learning curve for staff.Psychological resistance to change is compounded by a misconception that automation is an all-or-nothing proposition rather than a scalable process that can be integrated gradually and managed with already familiar tools.<h3>Deficiency of In-House Expertise</h3>Another barrier to automation is the perceived lack of in-house expertise. Small to medium-sized manufacturers, in particular, may not have the resources to sustain a dedicated IT department capable of deploying and maintaining advanced automation systems. They might assume that without specialists to guide the automation journey, the likelihood of successful implementation diminishes, causing them to shy away from taking the initial steps, despite the availability of more user-friendly automation tools.<h3>Misconceptions about Cost Implications</h3>A misunderstanding of the associated costs amplifies the hesitation regarding automation. Manufacturers tend to focus on the immediate financial outlay, while overlooking or underestimating the potential return on investment, which includes long-term savings and efficiency gains. This short-sighted view of finances obscures the actual value of automation and its capacity to streamline operations and cut costs over time.<h2>The Role of 3D Scanning in Bridging the Gap in Automated Manufacturing</h2>Integrating 3D scanners can be a manufacturer’s first manageable step toward automation.<a href="https://www.creaform3d.com/en/metrology-solutions/creaform-product-overview" target="_blank">Metrology-grade 3D measurement solutions</a> are remarkable for their versatility. They can be used at various stages throughout the manufacturing process, including <a href="https://www.creaform3d.com/en/applications" target="_blank">product design and development, quality control and quality assurance, reverse engineering</a>, and directly on production lines.<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1715330464303-P1130434_20x13-1536x1049.webp" style="width: 100%px;" class="fr-fic fr-dib">During the preliminary phase, 3D scanners contribute to accurate and detailed design models, laying the groundwork for quality. As products move through the manufacturing cycle, 3D scanners can be used for on-the-fly assessments and quality checks, and can easily transition between manual, hands-on operations to semi-automated processes.<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1715330480633-P1130421-1536x1152.webp" style="width: 100%px;" class="fr-fic fr-dib">This adaptability gives manufacturers a tangible starting point for automation, allowing them to start small and scale up, without the need for a full automation overhaul.For small to medium enterprises (SMEs), where every investment is even more weighed against its return, <a href="https://www.creaform3d.com/en/news/creaform-making-automated-quality-control-accessible-new-versatile-and-intuitive-hybrid" target="_blank">3D scanning technologies can be a cost-effective solution</a> to step into automation without immediately adopting a fully automated system. This allows SMEs to gradually increase their automation capabilities in tandem with their growth and financial comfort level. By implementing 3D scanning, these companies can reap the benefits of automation, such as increased accuracy and speed in production, without the full-scale investment that larger, more complex robotic systems would require.<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1715330487717-SON02915-2-1536x1024.webp" style="width: 100%px;" class="fr-fic fr-dib">Perhaps the most compelling argument for incorporating 3D scanning technology is its inherent scalability. It is a future-proof investment that supports a company’s growth trajectory. As businesses evolve and market demands shift, 3D scanning systems can be scaled to match increased production needs or to extend capabilities into new product lines.<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1715330504644-cube-r_automated-3d-scanning-cmm_front_cabinet_fender-apron-1536x1024.webp" style="width: 100%px;" class="fr-fic fr-dib">In short, 3D scanners are dynamic devices that can be leveraged in subsequent phrases with additional automation solutions. For example, 3D scanners could be used with robots or cobots for&nbsp;<a href="https://www.creaform-engineering.com/en/expertise/engineering/industrial-automation" target="_blank">seamless industrial automation.</a><h2>Selecting the Right 3D Scanner</h2>When choosing 3D scanning equipment, it is important to consider a few key factors to ensure that the chosen solution can meet your business needs both now and in the future. These factors are as follows:<ol><li>Scalability: It is best to opt for a technology that offers both handheld and automatable options. This allows for an initial investment in handheld devices that can be scaled up to automated systems as needs evolve and budgets allow.</li></ol><ol start="2"><li>Versatility: The technology should handle complex geometries and various materials without requiring extensive preparation, making it suitable for a wide range of applications and ensuring its long-term usefulness.</li></ol><ol start="3"><li>Speed and accuracy: These are essential for maintaining productivity and quality. The chosen technology should offer rapid data acquisition and high precision to meet tight tolerances and quality standards.</li></ol><ol start="4"><li>Software compatibility: It is essential to look for solutions that work seamlessly with external metrology and simulation software, allowing for efficient data transfer and utilization throughout the manufacturing process.</li></ol><ol start="5"><li>Supplier expertise: When selecting a 3D scanning manufacturer, it is important to choose one with extensive knowledge in both 3D scanning hardware and software. This ensures access to optimized support events when tackling the most challenging production workflows. Additionally, you should inquire if the vendor has previously worked on automation projects with its 3D scanners or has <a href="https://www.creaform3d.com/en/metrology-solutions/products/r-series-automated-quality-control-solutions" target="_blank">dedicated 3D measurement solutions for automated processes, such as quality control</a>. It is also worth checking if the supplier offers <a href="https://www.creaform3d.com/en/metrology-solutions/technology-integration-services" target="_blank">integration services</a> of its technologies within production environments.</li></ol><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1715330524815-MetraSCAN-R-BLACK-HD_car-frame-inspection_color-map-1-2048x1365.webp" style="width: 100%px;" class="fr-fic fr-dib"><h2 id="isPasted">Conclusion: 3D Scanners Smooth the Transition to Automation</h2>For manufacturers navigating the pressures of modern markets, starting with a versatile and upgradable 3D scanning solution offers a strategic pathway.By selecting the right 3D measurement technology and implementation approach, businesses can enhance their product development and quality control processes, boost efficiency, mitigate initial upfront costs, and set the stage for future automation.A strategic investment in 3D scanning not only addresses manufacturers’ current needs but also lays the foundation for scalable growth, ensuring they can adapt and thrive with emerging automation technologies.

Explore how manufacturers can overcome the challenges of automation complexity, cost misconceptions, and skill gaps with strategic 3D scanning solutions.

Navigating the Shift to Automated Manufacturing with 3D Scanning

<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1715299294035-5712x4284.jpg" style="width: 100%px;" class="fr-fic fr-dib">The robot spontaneously changed its gait to cross a terrain with gaps. 2024 EPFL/BioRob-CC-BY-SA 4.0With the help of a form of machine learning called deep reinforcement learning (DRL), the EPFL robot notably learned to transition from trotting to pronking – a leaping, arch-backed gait used by animals like springbok and gazelles – to navigate a challenging terrain with gaps ranging from 14-30cm. The study, led by the&nbsp;<a href="https://www.epfl.ch/labs/biorob/" target="_blank">BioRobotics Laboratory</a> in EPFL’s School of Engineering, offers new insights into why and how such gait transitions occur in animals. “Previous research has introduced energy efficiency and musculoskeletal injury avoidance as the two main explanations for gait transitions. More recently, biologists have argued that stability on flat terrain could be more important. But animal and robotic experiments have shown that these hypotheses are not always valid, especially on uneven ground,” says PhD student Milad Shafiee, first author on a paper published in <a href="https://www.nature.com/articles/s41467-024-47443-w" target="_blank" rel="noopener">Nature Communications</a>.Shafiee and co-authors Guillaume Bellegarda and BioRobotics Lab head Auke Ijspeert were therefore interested in a new hypothesis for why gait transitions occur: viability, or fall avoidance. To test this hypothesis, they used DRL to train a quadruped robot to cross various terrains. On flat terrain, they found that different gaits showed different levels of robustness against random pushes, and that the robot switched from a walk to a trot to maintain viability, just as quadruped animals do when they accelerate. And when confronted with successive gaps in the experimental surface, the robot spontaneously switched from trotting to pronking to avoid falls. Moreover, viability was the only factor that was improved by such gait transitions.“We showed that on flat terrain and challenging discrete terrain, viability leads to the emergence of gait transitions, but that energy efficiency is not necessarily improved,” Shafiee explains. “It seems that energy efficiency, which was previously thought to be a driver of such transitions, may be more of a consequence. When an animal is navigating challenging terrain, it’s likely that its first priority is not falling, followed by energy efficiency.”<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1715299423668-ezgif-3-e8b7f9ed78.jpg" style="width: 100%px;" class="fr-fic fr-dib">Guillaume Bellegarda, Milad Shafiee and Auke Ijspeert. 2024 EPFL/Jamani Caillet - CC-BY-SA 4.0<h2>A bio-inspired learning architecture</h2>To model locomotion control in their robot, the researchers considered the three interacting elements that drive animal movement: the brain, the spinal cord, and sensory feedback from the body. They used DRL to train a neural network to imitate the spinal cord’s transmission of brain signals to the body as the robot crossed an experimental terrain. Then, the team assigned different weights to three possible learning goals: energy efficiency, force reduction, and viability. A series of computer simulations revealed that of these three goals, viability was the only one that prompted the robot to automatically – without instruction from the scientists – change its gait.The team emphasizes that these observations represent the first learning-based locomotion framework in which gait transitions emerge spontaneously during the learning process, as well as the most dynamic crossing of such large consecutive gaps for a quadrupedal robot.“Our bio-inspired learning architecture demonstrated state-of-the-art quadruped robot agility on the challenging terrain,” Shafiee says.The researchers aim to expand on their work with additional experiments that place different types of robots in a wider variety of challenging environments. In addition to further elucidating animal locomotion, they hope that ultimately, their work will enable the more widespread use of robots for biological research, reducing reliance on animal models and the associated ethics concerns.<hr>ReferencesShafiee, M., Bellegarda, G. &amp; Ijspeert, A. Viability leads to the emergence of gait transitions in learning agile quadrupedal locomotion on challenging terrains. Nat Commun 15, 3073 (2024). <a href="https://doi.org/10.1038/s41467-024-47443-w" target="_blank" rel="noopener">https://doi.org/10.1038/s41467-024-47443-w</a>

Guillaume Bellegarda, Milad Shafiee and Auke Ijspeert. 2024 EPFL/Jamani Caillet - CC-BY-SA 4.0

A four-legged robot trained with machine learning by EPFL researchers has learned to avoid falls by spontaneously switching between walking, trotting, and pronking – a milestone for roboticists as well as biologists interested in animal locomotion.

Trotting robots reveal emergence of animal gait transitions

InDro Robotics is pleased to unveil details of its highly capable new R&amp;D drone.Running the Robot Operating System (ROS) and with powerful onboard compute capabilities, the drone is perfect for advanced Research and Development.“It’s a drone geared toward R&amp;D first and foremost,” explains Luke Corbeth, Head of R&amp;D Sales. “It truly is a flying robot – and you can program and use it in a very similar fashion to all our other robots.”There’s a real demand in the research world for open-source drones that can be programmed and run highly complex algorithms. These kinds of drones can be used to study swarm behaviour, object detection and identification, mapping in GPS-denied locations and much more.For some researchers, the budget go-to has been the <a href="https://www.bitcraze.io/products/crazyflie-2-1/" target="_blank">Crazyflie</a>, a micro-drone that uses a Raspberry Pi for compute. Its advantage is that it’s quite affordable. But its low cost, 27 gram weight and relatively low computing power means it has limitations – including the ability add sensors of any weight.“This drone can do so much more,” says Corbeth. “With the <a href="https://www.nvidia.com/en-us/autonomous-machines/embedded-systems/jetson-xavier-series/" target="_blank">NVIDIA Xavier NX</a> onboard for compute, it can effectively map entire environments. And when it comes to landing and object recognition, it’s truly phenomenal. It can even land on a moving vehicle.”Below: A look at InDro’s new drone, which comes complete with LiDAR, a depth-perception camera, 5G connectivity – and much more.<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1715267903013-20240403_192314.jpg" style="width: 100%px;" class="fr-fic fr-dib"><h3 id="isPasted">The Back Story</h3>If you’ve been following the latest news from InDro, you’ll be aware we have an incubation agreement with <a href="https://indrorobotics.ca/cypher-robotics-announces-captis-a-new-solution-for-industrial-scanning/" target="_blank">Cypher Robotics</a>. That company builds solutions for cycle counting and precision scanning in the industrial/supply chain space. InDro assisted with the development of its signature product, Captis.Captis integrates an autonomous ground robot with a tethered drone. As the Captis robot autonomously navigates even narrow stock aisles, the drone ascends from a tether attached to that ground robot. The drone then scans the barcodes (it’s code-agnostic) of the products on the shelves. All of that data is transferred seamlessly, in real-time, to the client’s Warehouse Management System (WMS), WCS (Warehouse Control System) and WES (Warehouse Execution System) software.The capabilities of Captis led to a partnership with global AI fulfilment experts <a href="https://www.greyorange.com/" target="_blank">GreyOrange</a> and leading global telco innovator <a href="https://www.ericsson.com/en" target="_blank">Ericsson</a>. The product debuted at the recent MODEX2024 conference (one of the biggies in the automated supply chain world), where it <a href="https://indrorobotics.ca/cypher-robotics-captis-system-generates-buzz-at-modex2024/" target="_blank">gained a *lot* of attention</a>.While working on the project, it was always clear the drone – thanks to multiple modifications – would be highly suitable as a research and development tool. It’s capable of machine vision/object recognition, machine learning, and can find its way around in completely unfamiliar, GPS-denied environments.“In fact, I have one client that’s using it for research in mines,” says Corbeth.&nbsp;<h3>The Jetson Difference</h3>NVIDIA has made quite a name for itself – and quite a profit for its shareholders – with its powerful AI-capable processors. The Jetson Xavier NX features a 6-core NVIDIA Carmel Arm®v8.2 64-bit processor running at speeds of up to 1.9 GHz. Its graphics processor unit features a 384-core NVIDIA Volta™ architecture with 48 Tensor Cores. Put it all together, and the computing power is astonishing: The Xavier NX is rated with a maximum achievable output of 21 TOPS – trillionoperations per second. (We were going to try to count, but thought it more efficient to rely on <a href="https://www.nvidia.com/en-us/autonomous-machines/embedded-systems/jetson-xavier-series/" target="_blank">NVIDIA’s specs</a> for this.)The LiDAR unit currently shipping with the drone also has some flex. It’s the Ouster 32-channel OS1 (Rev6.2). With a maximum range of 200 metres (90 metres on a dark, 10 per cent target), its powerful L3 chip is capable of processing scans of up to 5.2 million points per second with 128 channels of vertical resolution (again, we didn’t count). Hostile environment? No problem. The LiDAR can operate from -40°C to 60°C and has an IP68 Ingress Protection rating.“The OS1 is designed for all-weather environments and use in industrial automation, autonomous vehicles, mapping, smart infrastructure, and robotics,” <a href="https://ouster.com/products/hardware/os1-lidar-sensor" target="_blank">states its manufacturer</a>.&nbsp;“The OS1 offers clean, dense data across its entire field of view for accurate perception and crisp detail in industrial, automotive, robotics, and mapping applications.”The unit uses open source ROS and C++ drivers, and comes with Ouster’s Software Development Kit. Its ability to accurately sense its environment (down to distances of 0.5 metres away), combined with the NVIDIA processor and the depth camera also allows this machine to do something pretty extraordinary: It can recognise and land on a moving platform.“That’s a very challenging problem to solve and requires not only specific sensing but also really powerful onboard compute. This drone can do it,” explains Corbeth.Already, word about the product has been spreading. A number of units have already been sold to academic institutions for research purposes – and the team has been hard at work building and testing for the next set of orders (as seen below).<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1715268006509-shared-image.jpeg.webp" style="width: 100%px;" class="fr-fic fr-dib"><h3 id="isPasted">The Forge Connection</h3>Like all new products, the new drone required custom parts. We looked no further than <a href="https://indrorobotics.ca/indro-forge-prepped-for-next-phase-expansion/" target="_blank">InDro Forge</a>, our rapid prototyping and limited production run facility in Ottawa.Using state of the art additive and subtractive tools, the Forge team created custom mounts using carbon fibre and other strong but lightweight materials, while also ensuring the frame was robust enough to take on even the most challenging environments where these drones will be deployed.“InDro Forge has been critical to the finished product,” says Corbeth. “We wanted a look, feel and quality that matches this drone’s capabilities – and InDro Forge delivered.”<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1715268106286-R%26D+Drone+Custom+Parts.jpeg" style="width: 100%px;" class="fr-fic fr-dib"><h3 id="isPasted">InDro's Take</h3>We’re obviously excited about the capabilities of this new drone, and we’re not alone. Interest in this product from researchers has already been significant. In fact, we’re not aware of any other drone on the market offering this combination of specific capabilities.It was that void – in concert with our partnership with Cypher Robotics – that led to its creation.“InDro has always placed a great emphasis on the development of innovative new products,” says CEO Philip Reece. “We build new products at the request of clients and also develop our own when we see a market opportunity. In this case, the requirements for Cypher Robotics dovetailed nicely with demand for such a drone from researchers.”Production of the new drone is moving at a swift pace. If you’re interested in a briefing or demo, you can contact us <a href="https://indrorobotics.ca/connectwithanexpert/" target="_blank">here.</a>

An R&D Powerhouse

Unveiling the ROS-Based Drone

Robot Operating System (ROS) drivers were developed on Analog Devices products so that they can be readily used within a ROS ecosystem. This article will give an overview on how to use and integrate these drivers in their applications, products, and systems.

Enabling Robot Operating Systems - Introducing the ADI Trinamic Motor Controller ROS1 Driver

Pipe crawler conducting an inspection inside a wind turbine blade.

Inspecting turbine blades is critical to plan maintenance and avoid downtime. Empower your team to conduct quick, safe, detailed inspections without contracting.

Performing Remote Visual Inspections of Wind Turbine Blades

<section id="isPasted">In order to deploy mobile robots, self-driving cars, or drones into the real world, they need to be able to observe and understand the world around them, just like humans do. PhD candidate at TU/e Daan de Geus developed algorithms for automatic image recognition that are faster and more accurate than existing models. Last Wednesday, he obtained his PhD (with honors) at the Department of Electrical Engineering.</section><section tabindex="-1">De Geus thumbs through his dissertation and shows us a photo of a street, featuring different objects such as people, vehicles, traffic posts, and traffic lights (see image below). “Mobile robots and self-driving cars need to know what’s happening around them,” he says.Among other things, they need to be able to recognize and localize objects, so they can take these into account. “This allows them to move around or toward a certain object, and maybe pick it up to carry out a task.”<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1714637661762-csm_frankfurt_000001_030669_images_0a92c1e725.jpg" style="width: 100%px;" class="fr-fic fr-dib">Traffic in Frankfurt. Image: Daan de GeusTo make robots aware of their environment, different computer vision techniques are used. It concerns an area of research that revolves around automatic extraction of relevant information from camera footage. “Simply put, we try to make models that can distill as much information from a photo as possible,” the PhD candidate explains.“The goal of computer vision is to make a system that mimics our own visual system, allowing computers to see in the same way humans do and, as a result, interact properly with the world around them.”His research focuses on improving image recognition techniques in the area of scene understanding. This is a small, but crucial part of computer vision. “Its goal is to recognize different objects and regions in an image and to give them a semantic label, such as ‘streetlight’, ‘road’, ‘car’, or ‘person’,” he explains.In other words, the objects are given a significance that is clear to humans.<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1714637694452-csm_scene_understanding_example_189e7cc585.jpg" style="width: 100%px;" class="fr-fic fr-dib">An example of scene understanding. Objects (person, traffic light, car) and image regions (the road) are automatically identified, color-coded, and semantically labeled. Image: Daan de Geus<h2 id="isPasted">Accuracy and efficiency</h2>For automatic image analyses, neural networks are used – systems that learn to carry out a certain task by receiving training involving a large amount of data. By training these neural networks in a focused manner, you end up with different models that specialize in specific tasks, such as identifying all cars in an image.In the first part of his dissertation, De Geus looked at how you can improve the accuracy and efficiency of those models. “Improving one aspect is often at the expense of another. To get more accurate results, existing methods tend to lose some of their efficiency, and vice versa,” he explains.“Which makes sense, because greater accuracy often demands more computing power, which directly impacts efficiency.” The big question, therefore, is: how can you improve these aspects without having to compromise?He came up with several solutions to this conundrum. “Efficiency is important for two reasons. You want the algorithm to use as little energy as possible and to make a prediction as quickly as possible,” he says. Speed is of the essence in self-driving vehicles, as cars must be able to respond to situations that arise in a timely manner.“If the calculation takes a second, it may already be too late to take action.”Using a process known as model unification, he combined two models to create a more efficient one. “Some tasks focus on the objects in the foreground, such as cars and people; others on ‘background regions’ such as vegetation and the sky. These tasks are performed by two different neural network modules, as each module specializes in a different task,” he explains.“Using two network modules isn’t very efficient, because you have to run them parallel to each other.” He found out that you can supply extra information to the network module processing the background information, allowing it to also identify foreground objects.“As a result, the module for foreground objects is no longer necessary, which greatly enhances efficiency. This makes this model twice as fast as its precursors, while its accuracy is comparable.”<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1714637727049-csm_image_with_patches_0ce2b38ff7.jpg" style="width: 100%px;" class="fr-fic fr-dib">Image with image regions. Image: Daan de GeusAnother way of improving efficiency is based on the observation that many regions in an image are very similar. For instance, the sky generally takes up a large part of any picture you take outside. “In spite of the fact that a lot of information is similar, neural networks process each image region separately, which is very inefficient,” says De Geus.The entire image is divided up into patches consisting of pixels (see image below). Normally, these would be labeled and assessed individually, but the PhD candidate developed a method for clustering patches with comparable information, reducing the total number of patches and, by extension, requiring less computing power.“This allows us to improve speed by as much as 110%, without taking away from the accuracy.”Incidentally, this method for grouping similar image regions is widely applicable to many different models and a great many applications other than self-driving cars and mobile robots. For example, you could use these algorithms to segment medical images. “Pretty much in any situation where you need to automatically analyze images.”<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1714637752641-csm_token_sharing_diagram_c977af3012.jpg" style="width: 100%px;" class="fr-fic fr-dib">To improve efficiency, semantically comparable patches can be clustered. Image: Daan de Geus<h3 id="isPasted">Multiple abstraction levels</h3>In the second part of his dissertation, he focused on what he calls multiple abstraction levels. “Existing algorithms either focus on entire objects, such as cars, or on their components, such as car tires or the license plate,” he explains.His goals was to develop an algorithm that can simultaneously understand an image at multiple abstraction levels. “This will allow a mobile robot to observe both the car and its components, as well as other objects and the background, all at the same time, which gives it a comprehensive picture of the surroundings,” says De Geus.<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1714637778377-csm_frankfurt_000001_030669_panoptic_5abb4ba438.jpg" style="width: 100%px;" class="fr-fic fr-dib">Automatic recognition of objects and image regions. Image: Daan de GeusYou can have both algorithms carry out their work separately and merge the results afterwards, but this is cumbersome. And it can also lead to conflicts between the two calculations. Instead, he developed a new algorithm that can identify objects and components all at once.“That’s not only more accurate, but also more efficient.” What’s more, this improved model is widely applicable and has many advantages. “It means, for instance, that a robot isn’t just able to recognize a door, but also the door handle, and to understand that the latter is part of the door. Which, in turn, allows it to open the door,” he says to illustrate.<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1714637801563-csm_frankfurt_000001_030669_pps_3b32a5d47d.jpg" style="width: 100%px;" class="fr-fic fr-dib">A more detailed segmentation, where components of objects are also identified. Image: Daan de GeusTo be able to use self-driving cares or mobile robots in practice, there’s a lot left to do still. “Image recognition is just a tiny part of the puzzle,” he admits.&nbsp;And there are plenty of remaining challenges in the area. For one thing, models could be developed that can recognize more objects at once and that operate even faster and more accurately.The generalization of the model could also be improved, so it also produces good results when the data are different from the data sets with which the network was trained, allowing it to perform better in real-world scenarios.But at least his dissertation constitutes a step forward in making the system more capable, accurate, and efficient.<hr>Source: <a href="https://www.cursor.tue.nl/en/news/2024/april/week-3/home-stretch-teaching-robots-to-see-better/" target="_blank" rel="noreferrer">Cursor/Martina Silbrníková</a>Daan de Geusdefended his thesis“<a href="https://research.tue.nl/en/publications/characterization-of-abdominal-aortic-aneurysms-using-time-resolve" target="_blank" rel="noreferrer">Advances in Scene Understanding - Towards Efficient Image Segmentation at Multiple Abstraction Levels</a>” at the Department of Electrical Engineering on April 17th, 2024.Supervisors: Gijs Dubbelman en Peter de With</section>

In his doctoral research, Daan de Geus worked on advanced image processing methods that allow robots and cars to better recognize what they see around them.

Teaching robots to see better

This comprehensive article dives deep into the world of robotics, exploring the history, types, engineering components, applications, and future trends of robots, offering readers an in-depth understanding of how these remarkable machines work and shape our lives.

What is Robotics? A Comprehensive Guide to its Engineering Principles and Applications

Spacebit-Fri Jun 30 2023 23:59:59 GMT+0200 (Central European Summer Time)

Spacebit-Mon Jul 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

Spacebit-Thu Aug 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

Spacebit-Thu Aug 31 2023 23:59:59 GMT+0200 (Central European Summer Time)

Spacebit-Sat Sep 30 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

RAD Lab-Fri Jun 30 2023 23:59:59 GMT+0200 (Central European Summer Time)

RAD Lab-Mon Jul 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

RAD Lab-Thu Aug 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

RAD Lab-Thu Aug 31 2023 23:59:59 GMT+0200 (Central European Summer Time)

RAD Lab-Sat Sep 30 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

NAVER LABS Europe-Fri Jun 30 2023 23:59:59 GMT+0200 (Central European Summer Time)

NAVER LABS Europe-Mon Jul 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

NAVER LABS Europe-Thu Aug 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

NAVER LABS Europe-Thu Aug 31 2023 23:59:59 GMT+0200 (Central European Summer Time)

NAVER LABS Europe-Sat Sep 30 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

Franka Emika-Fri Jun 30 2023 23:59:59 GMT+0200 (Central European Summer Time)

Franka Emika-Mon Jul 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

Franka Emika-Thu Aug 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

Franka Emika-Thu Aug 31 2023 23:59:59 GMT+0200 (Central European Summer Time)

Franka Emika-Sat Sep 30 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

PEK AUTOMOTIVE-Fri Jun 30 2023 23:59:59 GMT+0200 (Central European Summer Time)

PEK AUTOMOTIVE-Mon Jul 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

PEK AUTOMOTIVE-Thu Aug 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

PEK AUTOMOTIVE-Thu Aug 31 2023 23:59:59 GMT+0200 (Central European Summer Time)

PEK AUTOMOTIVE-Sat Sep 30 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

READY Robotics-Fri Jun 30 2023 23:59:59 GMT+0200 (Central European Summer Time)

READY Robotics-Mon Jul 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

READY Robotics-Thu Aug 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

READY Robotics-Thu Aug 31 2023 23:59:59 GMT+0200 (Central European Summer Time)

READY Robotics-Sat Sep 30 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

NEURA Robotics-Fri Jun 30 2023 23:59:59 GMT+0200 (Central European Summer Time)

NEURA Robotics-Mon Jul 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

NEURA Robotics-Thu Aug 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

NEURA Robotics-Thu Aug 31 2023 23:59:59 GMT+0200 (Central European Summer Time)

NEURA Robotics-Sat Sep 30 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

Blueprint Lab-Fri Jun 30 2023 23:59:59 GMT+0200 (Central European Summer Time)

Blueprint Lab-Mon Jul 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

Blueprint Lab-Thu Aug 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

Blueprint Lab-Thu Aug 31 2023 23:59:59 GMT+0200 (Central European Summer Time)

Blueprint Lab-Sat Sep 30 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

Adapta Robotics-Fri Jun 30 2023 23:59:59 GMT+0200 (Central European Summer Time)

Adapta Robotics-Mon Jul 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

Adapta Robotics-Thu Aug 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

Adapta Robotics-Thu Aug 31 2023 23:59:59 GMT+0200 (Central European Summer Time)

Adapta Robotics-Sat Sep 30 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

Soft Robotics Podcast-Fri Jun 30 2023 23:59:59 GMT+0200 (Central European Summer Time)

Soft Robotics Podcast-Mon Jul 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

Soft Robotics Podcast-Thu Aug 31 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

Soft Robotics Podcast-Thu Aug 31 2023 23:59:59 GMT+0200 (Central European Summer Time)

Soft Robotics Podcast-Sat Sep 30 2023 23:59:59 GMT+0300 (Eastern European Summer Time)

Introduction

Edge AI Market Analysis and Trends

Healthcare and Medical Applications

Industrial IoT and Manufacturing

Smart Cities and Urban Infrastructure

Retail and Customer Experience

Energy Efficiency and Sustainability

Agriculture and Food Production 

Automotive and Transportation

Generative AI at the Edge

Edge AI Challenges and Real-World Mitigations

The Future of Edge AI

Exploring the Dynamic World of Edge AI Applications Across Industries


2024 State of Edge AI Report

robotics

A pick and place robot in an industrial setup

Understanding the inception, evolution, composition, and types of pick and place robots and their impact on various industrial sectors 

Pick and Place Robots: An In-Depth Guide to Their Functionality and Applications

Surface roughness can be measured with a profilometer

Measuring surface roughness allows metrologists to establish the roughness of a surface for quality control or other purposes. But how exactly is it done?

Measuring Surface Roughness: A Comprehensive Guide

An AI-generated depiction of a modern day chipset

Gaining a comprehensive understanding of the distinct capabilities and applications of TPUs and GPUs is crucial for developers and researchers aiming to navigate the complex and rapidly changing terrain of artificial intelligence.

TPU vs GPU in AI: A Comprehensive Guide to Their Roles and Impact on Artificial Intelligence

Optimizing Operational Excellence: Synergies of Human-Machine Interfaces, Motion Control, IIoT, and Robust Power Supply in Industrial Automation

Understanding Industrial Automation: A Comprehensive Guide

Unlocking the Potential of Zener Diodes: A Deep Dive into Principles and Practical Applications

Zener Diode: A Comprehensive Guide to Its Principles and Applications

Robotics

2024 State of Edge AI Report

Introduction

1. Edge AI Market Analysis and Trends

2. Healthcare and Medical Applications

3. Industrial IoT and Manufacturing

4. Smart Cities and Urban Infrastructure

5. Retail and Customer Experience

6. Energy Efficiency and Sustainability

7. Agriculture and Food Production

8. Automotive and Transportation

9. Generative AI at the Edge

10. Edge AI Challenges and Real-World Mitigations

11. The Future of Edge AI

What is Robotics? A Comprehensive Guide to its Engineering Principles and Applications

Profiles