<p dir="ltr" id="isPasted">Training robots to learn new skills has been a complex process restricting the work for programmers and roboticists. Scientists who study human-robot interaction often focus on easing the robot training process for it to learn new tasks in an unstructured environment. Today, most of the techniques for teaching robots are limited to the use of remote hardware or imitating pre-recorded demonstrations. The idea of teaching robots through a self-learning mechanism by watching humans do the task— allows the deployment of these robots in a new setting. As fancy as it sounds, there are certain challenges faced by the vision-based learning process as there exists a clear embodiment gap between the human demonstrator and the robotic hardware that executes the task.&nbsp;</p><p dir="ltr">As reported by the existing research [1,2], even though humans and robots are capable of performing the same task, the differences in the physique can make easy tasks complex or even impossible to do. A team of researchers from Stanford University, Robotics at Google and the University of California at Berkeley collaborated to apply well-established reporting of human-robot interaction to focus on teaching robots from third-person observations of demonstrations [3]. The researchers identified the opportunity to investigate the scope of self-supervised methods for vision-based learning techniques while considering the differences in human-robot embodiments, like shape, action and end-effector dynamics.&nbsp;</p><h2 dir="ltr">Robots See, Robots Do</h2><p dir="ltr">In the paper, “XIRL: Cross-embodiment Inverse Reinforcement Learning,” the researchers carried out work to solve the challenges of embodiment through self-supervised mechanisms of high-level task objectives from videos. Rather than focusing on the way humans perform tasks and relating it to robot training, the proposition summarizes the knowledge in the form of a reward function that is invariant to the embodiment differences. This means it is now possible to learn vision-based reward functions from cross-embodiment demonstration videos that are robust to the differences mentioned earlier.&nbsp;</p><p dir="ltr"><span class="fr-img-caption fr-fic fr-dib" style="width: 766px;"><span class="fr-img-wrap"><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1646257426964-Cross-Embodiment+Inverse+Reinforcement+Learning.gif" style="width: 800px;" class="fr-fic fr-dib" alt="Cross-Embodiment Inverse Reinforcement Learning"><span class="fr-inner" spellcheck="false">XIRL self-supervises reward functions using temporal cycle consistency (TCC), then uses them for reinforcement learning to learn new skills from third-person demonstrations [Image Source: Research Paper]</span></span></span></p><div style='pointer-events: none; overflow: hidden; padding-bottom: 2px; z-index: auto; position: absolute; left: 0px; top: 429.75px; height: 63px; width: 766px; font-style: normal; font-variant: normal; font-weight: 400; font-stretch: 100%; font-size: 14px; font-family: "PT Serif", serif; text-align: center; text-transform: none; letter-spacing: normal; word-spacing: 0px; tab-size: 8;'><br></div><p>“The learned rewards can be used together with reinforcement learning (RL) to teach the task to agents with new physical embodiments through trial and error,” the team explains in a <a href="https://ai.googleblog.com/2022/02/robot-see-robot-do.html" rel="noopener noreferrer" target="_blank">Google AI blog post</a>. “Our approach is general and scales autonomously with data — the more embodiment diversity presented in the videos, the more invariant and robust the reward functions become.”</p><p dir="ltr">Even in the presence of embodiment differences, there still exist visual cues that indicate the task objective. The fundamental idea behind the XIRL is to discover the key moments in the videos of different lengths and clusters to encode task progression. The work carried out in XIRL leverages the Temporal Cycle Consistency (TCC) [4] that aligns the video accurately while learning the visual cues for understanding the task objective without requiring any ground-truth correspondences. The reward function in the learning process is the negative Euclidean distance between the current observation and the goal observation in the learned embedding space. This reward function is then fed into a standard Markov decision process and uses a reinforcement learning algorithm to learn the third-person demonstrated behavior.&nbsp;</p><h2 dir="ltr">Transfer policies from humans to robots in simulation</h2><p dir="ltr">Along with the cross-embodiment inverse reinforcement learning label-free framework, the team proposes a cross-embodiment imitation learning benchmark, X-MAGICAL that features several tried agents with different embodiments performing the same defined task including one thousand expert demonstrations for a single agent. “The goal of X-MAGICAL is to test how well imitation or reward learning techniques can adapt to systematic embodiment gaps between the demonstrator and the learner,” the team notes. “These differences in execution speeds and state-action trajectories pose challenges for current LfD techniques, and the ability to generalize across embodiments is precisely what this benchmark evaluates.</p><p dir="ltr"><span class="fr-img-caption fr-fic fr-dib" style="width: 766px;"><span class="fr-img-wrap"><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1646257571549-Comparison+of+XIRL+with+baselines+using+the+simulated+State+Pusher+environment.png" style="width: 800px;" class="fr-fic fr-dib" alt="Comparison of XIRL with baselines using the simulated State Pusher environment"><span class="fr-inner" spellcheck="false">Comparison of XIRL with baselines using the simulated State Pusher environment [Image Source: Research Paper]</span></span></span></p><div style='pointer-events: none; overflow: hidden; padding-bottom: 2px; z-index: auto; position: absolute; left: 0px; top: 477.5px; height: 39px; width: 766px; font-style: normal; font-variant: normal; font-weight: 400; font-stretch: 100%; font-size: 14px; font-family: "PT Serif", serif; text-align: center; text-transform: none; letter-spacing: normal; word-spacing: 0px; tab-size: 8;'><br></div><p>The results show that the XIRL significantly outperforms the alternative methods on both the X-MAGICAL benchmarks and the human-to-robot transfer benchmark. The team demonstrates the effectiveness of the proposed methodology by using third-person demonstrations on the State Pusher [5] setting where the reward function is first learned on real-world human demonstration and then used to teach a Sawyer arm to perform the task in simulation. The XIRL methodology trained on real-world demonstration (purple) reaches 80% of the total performance around 85% faster than the reinforcement learning with videos (RLV) as baseline (orange).</p><h2 dir="ltr">Conclusion on XIRL be used for learning RL policies</h2><p dir="ltr">In the proposed approach to tackle the cross-embodiment imitation problems, the XIRL learns an embodiment invariant reward function technique using a temporal cycle consistency objective. Compared to baseline alternatives, XIRL reward functions are significantly more efficient. While the experimental results are quite promising, the team plans to show policy learning on a real robot post-COVID pandemic.&nbsp;</p><p dir="ltr">This research article was published on Cornell University’s research sharing platform, arXiv under <a href="https://arxiv.org/pdf/2106.03911.pdf" rel="noopener noreferrer" target="_blank">open access terms</a>.</p><h2 dir="ltr">References</h2><p dir="ltr">[1] Abhishek Gupta, Coline Devin, YuXuan Liu, Pieter Abbeel, Sergey Levine: <em>Learning Invariant Feature Spaces to Transfer Skills with Reinforcement Learning</em>. DOI <a href="https://arxiv.org/abs/1703.02949" target="_blank">arXiv:1703.02949 [cs.AI]</a>.</p><p dir="ltr">[2] Pierre Sermanet, Corey Lynch, Yevgen Chebotar, Jasmine Hsu, Eric Jang, Stefan Schaal, Sergey Levine: <em>Time-Contrastive Networks: Self-Supervised Learning from Video</em>. DOI <a href="https://arxiv.org/abs/1704.06888v3" target="_blank">arXiv:1704.06888 [cs.CV]</a>.</p><p dir="ltr">[3] Kevin Zakka, Andy Zeng, Pete Florence, Jonathan Tompson, Jeannette Bohg, Debidatta Dwibedi: <em>XIRL: Cross-embodiment Inverse Reinforcement Learning</em>. DOI <a href="https://arxiv.org/abs/2106.03911" target="_blank">arXiv:2106.03911 [cs.RO]</a>.</p><p dir="ltr">[4] D. Dwibedi, Y. Aytar, J. Tompson, P. Sermanet, and A. Zisserman. Temporal cycle-consistency learning. In The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2019.</p><p dir="ltr">[5] K. Schmeckpeper, O. Rybkin, K. Daniilidis, S. Levine, and C. Finn. Reinforcement learning with videos: Combining offline observations with interaction, 2020</p>

Real-world-demo cross-embodiment setting [Image Source: Research Paper]

Human-robot interaction focuses on teaching robots from third-person observations of demonstrations to make it invariant and robust to the embodiment differences.

Vision-based learning technique for robots during training

<p dir="ltr" id="isPasted">As technological advances come through, humans and robots are working more closely together to increase the productivity of industries and the quality of manufactured products, resulting in improved efficiency and growth. The research interest is to design humanoid robots that can become co-workers more than mere tools. From industrial robots to domestic and service robots, roboticists and computer engineers are witnessing the need for human-robot collaboration. For deploying robotic systems to work alongside human resources, it is important for humans to trust the abilities of their functionalities and willingness to use the robot as a co-worker.&nbsp;</p><p dir="ltr">Human-robot interaction has been widely studied in the past decade, evaluating robotic failures and varying severity in terms of impactful consequences. However, the existing research evaluation was restricted to artificial settings which lacked real-life external parameters, resulting in an inaccurate response. Personal relevance (PeR), defined as the level of involvement with an object (in this case, Robot) has been impacting the user preferences and perceptions of the robots. Celsi and Olson found that PeR affects human perception [1] which is why a group of scientists from the Ben-Gurion University of the Negev, in Israel [2] speculated that this will affect robotic failures’ perception.&nbsp;</p><h2 dir="ltr">The Evaluation</h2><p dir="ltr">In the research article, “Is it personal? The impact of personally relevant robotic failures (PeRFs) on humans' trust, likeability, and willingness to use the robot,” the team carried out a study to evaluate the factors that affect how humans trust the robots– without affecting the productivity. The paper primarily focuses on the impact of personally relevant robotic failures (PeRFs) which would reduce the trust in robots and robots’ likeability and willingness to use more than failures that are not personal to the user.&nbsp;</p><p dir="ltr">While the existing research failed to evaluate the factors in real-life scenarios, the team carried out a series of three experiments in the experimental environment, designed to simulate a realistic laundry sorting setup. To investigate the role of PeR in human-robot interaction, the three different laboratory experiments included one with damaging to property, the second with causing financial losses, followed by the first-person versus third-person failure scenarios. For the collaborative task of laundry sorting, a total of 132 participants engaged with the robot with the results indicating the impact of PeRFs on perceptions of the robot differed across the experimental setups.&nbsp;</p><p dir="ltr"><span class="fr-img-caption fr-fic fr-dib" style="width: 766px;"><span class="fr-img-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjQ1Njg4MDQ4OTY2LVJvYm90aWMtSHVtYW4gSW50ZXJhY3Rpb24ucG5nIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" style="width: 800px;" class="fr-fic fr-dib" alt="Summary of Experiment 1 to check user preference after robotic failures"><span class="fr-inner" sapling-id="c0457644-9444-41d7-8754-bd0a3c2e7bd3" sapling-input-id="386deafc-a898-4a2f-957a-5553bf99db65" spellcheck="false">Experiment 1: The interaction between the level of personal relevance and the session in terms of trust and willingness to use. [Image Credit: Research Paper]</span></span></span></p><div is-sapling-overlay="true" style="pointer-events: none; overflow: hidden; padding-bottom: 2px; z-index: auto; position: absolute; left: 0px; top: 477.7px; height: 63px; width: 766px; font-style: normal; font-variant: normal; font-weight: 400; font-stretch: 100%; font-size: 14px; font-family: &quot;PT Serif&quot;, serif; text-align: center; text-transform: none; letter-spacing: normal; word-spacing: 0px; tab-size: 8;"></div><p></p><div style='pointer-events: none; overflow: hidden; padding-bottom: 2px; z-index: auto; position: absolute; left: 0px; top: 477.5px; height: 63px; width: 766px; font-style: normal; font-variant: normal; font-weight: 400; font-stretch: 100%; font-size: 14px; font-family: "PT Serif", serif; text-align: center; text-transform: none; letter-spacing: normal; word-spacing: 0px; tab-size: 8;'><br></div><p>Specifically, in the first experiment, the team aims to evaluate the impact of a failure that may cause damage to participants and examine the interaction between PeR and perceived failure severity. The severity differed from being high when the robot dropped a clothing object into the trash can instead of the laundry bin while the severity dropped when the robot dropped the clothing item onto the floor. The second experiment of financial loss due to delays caused by robotic failures reported that the failures with PeR can negatively impact humans’ trust in robots. The results from experiment two suggest that the perceived trust before failure was more than trust after failure, indicating that the failures decreased the user’s trust.&nbsp;</p><p dir="ltr">Finally, in the last experimental setup, the failures were wrong gender identification of the participant or the experimenter. The conversation between the robot and the participant was in Hebrew, a gendered language that addresses the other person specifying the person’s gender. In high PeR, the robot incorrectly identified the participant’s gender while the lower PeR suggested that the robot incorrectly identified the experimenter’s gender. Surprising in this result, there was no difference between trust before failure and trust after failure, while no difference between LWtU before failure and after failure. This indicated that the manipulation of the interaction PeRF had no impact on the participant.&nbsp;</p><p dir="ltr">The findings of the research article indicate that the robot’s failure changed human perception and impacted the extent to which users are ready to trust a robot for its designed functionality. The willingness to use had been affected by the robot’s failure which sparked an interesting observation fueling future studies. This research article was published on Cornell University’s research sharing platform, arXiv under <a href="https://arxiv.org/abs/2201.05322" target="_blank">open access terms</a>. &nbsp;&nbsp;</p><h2 dir="ltr">Reference</h2><p dir="ltr">[1] R. L. Celsi and J. C. Olson, “The Role of Involvement in Attention and Comprehension Processes,” JOURNAL OF CONSUMER RESEARCH, vol. 15, pp. 210–224, 1988.</p><p dir="ltr">[2] Romi Gideoni, Shanee Honig, Tal Oron-Gilad: Is it personal? The impact of personally relevant robotic failures (PeRFs) on humans' trust, likeability, and willingness to use the robot. DOI <a href="https://arxiv.org/abs/2201.05322" target="_blank">arXiv:220105322</a> [cs.RO]</p>

The robotic-human interaction experimentation– Laundry Sorting Workstation [Image Credit: Research Paper]

Impact of personally relevant robotic failures (PeRFs) which would reduce trust in robots and robots’ likeability and willingness to use more than failures that are not personal to the user. 


Did personally relevant robotic failures affect human perception?

<p id="isPasted"><a href="https://sites.google.com/site/tapomayukh/home" target="_blank">Tapomayukh Bhattacharjee</a> has been working on assistive robotics for approximately a decade, and in speaking to the people who would benefit from new technologies, he’s come to a few realizations.</p><p>“I started chatting a lot with the stakeholders – people with mobility limitations, the caregivers who are helping them, and occupational therapists – and I felt like robotics can truly make a difference in their lives, when it’s ready to be deployed,” said Bhattacharjee, hired in July 2021 as an assistant professor of computer science in the Cornell Ann S. Bowers College of Computing and Information Science.</p><p><span class="fr-video fr-deletable fr-fvc fr-dvb fr-draggable" contenteditable="false" draggable="true"><iframe src="https://www.youtube.com/embed/f-NboRkoTDs?&amp;wmode=opaque&amp;rel=0" frameborder="0" allowfullscreen="" class="fr-draggable" style="width: 768px; height: 439px;"></iframe></span><em>Tapomayukh Bhattacharjee received a National Robotics Initiative collaborative grant for a project that aims to address the way people with mobility issues are given a chance for improved control and independence over their environments, especially how they are fed.&nbsp;</em></p><p id="isPasted">Among the basic activities of daily living that robotics could impact, Bhattacharjee is currently focused on eating. And a&nbsp;<a href="https://www.nsf.gov/awardsearch/showAward?AWD_ID=2132846&HistoricalAwards=false" target="_blank">four-year, $1.5 million grant</a> from the National Science Foundation’s National Robotics Initiative will help him and his&nbsp;<a href="https://emprise.cs.cornell.edu/" target="_blank">EmPRISE Lab</a> develop assistive robotics for people with physical disabilities and their caregivers.</p><p>“Feeding is one of the most basic activities,” he said. “Imagine yourself asking someone else to feed you every morsel of food in your day-to-day life. It just completely takes away the sense of independence. And so, if we could solve this feeding challenge, if a person could perceive this robot as an extension of their own body, then they will feel much more independent. That’s why I am so passionate about solving this.”</p><p>Bhattacharjee – who received his Ph.D. in 2017 from the Georgia Institute of Technology, and developed a prototype robot during his postdoctoral work at the University of Washington prior to joining the Cornell Bowers CIS faculty – fully understands the nature of the challenge in front of him.</p><p>“Despite great strides taken toward sustainable solutions in controlled environments,” he wrote in his grant proposal, “robots are far from ready for adoption in real-home environments as long-term caregiving solutions.”</p><p>To address the eating challenge, Bhattacharjee will focus on three key elements:</p><ul><li><strong>Bite acquisition:</strong> His team will derive algorithms that learn by obtaining human feedback – which they refer to as human-in-the-loop manipulation – to perform difficult tasks related to eating, particularly with different forms of food;</li><li><strong>Bite transfer:</strong> The goal is to develop safe control and learning algorithms for bite transfer – that is, the robot actually placing the food safely in a person’s mouth, particularly if the person has profound physical limitations and cannot move to take the food; and</li><li><strong>Feeding with user preferences:</strong> The team will develop active learning and shared autonomy algorithms that capture, over time, human preferences of bite transfer trajectories.</li></ul><p>“The main lessons we learned in our experiments [at Washington] was that every user is different,” he said. “Everybody has their personal preferences. And if we really want a long-term caregiving solution, the solution needs to be personalized to the user. And just like a patient and a caregiver need time to get used to each other, it’s the same with a patient and a robot.”</p><p>The main target population Bhattacharjee is focusing on with his assistive technology is people with spinal cord injuries – those who are limited physically but not necessarily cognitively, so they would be able to help train the robot.</p><p>One key element of the work will be leveraging the idea of human feedback – specifically by embracing the inevitable physical contact between the robot and human, such as when placing the utensil inside the mouth.</p><p>“If you think about robots in airports, self-driving cars, robots in grocery stores, it’s all about avoiding obstacles,” Bhattacharjee said. “But realistically, if I’m grabbing for something, my arm might touch something else, right? The whole point is, can we leverage this contact, while being safe and efficient at the same time?”</p><p>The grant period starts Jan. 1, 2022, and runs through 2025.</p>

Doctoral student Rajat Kumar Jenamani, a member of Tapomayukh Bhattacharjee’s EmPRISE Lab, prepares to take a piece of a banana from an assistive robot prototype. Ryan Young/Cornell University

Tapomayukh Bhattacharjee has been working on assistive robotics for approximately a decade, and in speaking to the people who would benefit from new technologies, he’s come to a few realizations.

Robot-assisted feeding the focus of 1.5M USD NSF grant

<p>YOLO is a non-anthropomorphic social robot designed to stimulate creativity in children. This robot was envisioned to be used by children during free-play where they use the robot as a character for the stories they create. During play, YOLO makes use of creativity techniques that promote the creation of new storylines.&nbsp;</p><p>Therefore, the robot serves as a tool that has the potential to stimulate creativity in children during the interaction. Particularly, YOLO can stimulate divergent and convergent thinking for story creations. Additionally, YOLO can have different personalities, providing it with socially intelligent and engaging behaviors.&nbsp;</p><p>This work provides open-source and open access to YOLO’s hardware. The design of the robot was guided by psychological theories and models on creativity, design research including user-centered design practices with children, and informed by experts working in the field of creativity.&nbsp;</p><p>Specifically, it relies on established theories of personality to inform the social behavior of the robot, and on theories of creativity to design creativity stimulating behaviors. The design decisions were then based on design fieldwork with children.&nbsp;</p><p>The end product is a robot that communicates using non-verbal expressive modalities (lights and movements) equipped with sensors that detect the playful behaviors of children. YOLO has the potential to be used as a research tool for academic studies, and as a toy for the community to engage in personal fabrication.&nbsp;</p><p>The overall benefit of this proposed hardware is that it is open-source, less expensive than existing ones, and one that children can build by themselves under expert supervision.</p><p><span class="fr-video fr-deletable fr-fvc fr-dvb fr-draggable" contenteditable="false" draggable="true"><iframe src="https://www.youtube.com/embed/e-K3J5UZ9M4?&amp;wmode=opaque&amp;rel=0" frameborder="0" allowfullscreen="" class="fr-draggable" style="width: 789px; height: 450px;"></iframe></span><br></p><p><strong>Design&nbsp;</strong></p><p>YOLO was designed using formative research and feasibility studies. Formative research consists of an exploratory research methodology whose goal is to guide a design process, allowing ongoing intermediate assessments. By identifying and solving concrete usability problems throughout the entire process design, current systems can be improved or new ones are developed.&nbsp;</p><p>Feasibility studies consist of are pieces of research performed before the main study, used to estimate crucial parameters that are needed to design the main study. These intermediary studies will help to prepare for full-scale and large research intervention, allowing for low-cost improvements before experimentally testing the effectiveness of interventions or products. Additionally, the Double-Diamond Design Process Model was used to design YOLO, an established theory in the field of design research.&nbsp;</p><p>The contextual design was used by involving children at different design stages of the robot. The importance of contextual design lies in the fact that it validates the already embodied system where the robot will exist, capturing the child’s world as the floor for design decisions. Data gathered during fieldwork was the base criterion for deciding what needs to be addressed, what the robot should do, and how it should be structured within the reality of children. Additionally, the Big Five Model of Personality informs the design of the social behaviors for the robot.&nbsp;</p><p>Finally, the robot was aimed to stimulate creativity during play. Therefore, the creativity-stimulating behavior of YOLO was based on two established creativity techniques for idea generation, an important stage in story creation. These techniques are called ‘‘contrast” and ‘‘mirror”. While the Contrast technique stimulates divergent thinking, the Mirror technique is responsible for the development of convergent thinking. Both modes of thinking are required to establish the emergence of creativity.&nbsp;</p><p>Therefore, YOLO’s behavior is not only loaded with social behaviors (derived from personality theories) but also with behaviors that can lead to creativity stimulation in children.</p><p><br></p><p><span class="fr-img-caption fr-fic fr-dib" style="width: 736px;"><span class="fr-img-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjE5NjQzODg5NDE1LTQuanBnIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" style="width: 736px;" class="fr-fic fr-dib"><span class="fr-inner" spellcheck="false">YOLO’s exploded view (on the left) and section analysis with component coloring (on the right)</span></span></span></p><p>The design files are in STL format and ready to be 3D printed. Some of these files can be converted to a DXF format, adding a laser cutting option for faster and cheaper opportunities to fabricate YOLO. If opting to laser cut some of the components, note that the thickness of the laser cutting material should correspond to the CAD model dimensions.</p><p><br></p><ul><li>Shell– File with the cover of the robot. This is the largest 3D printing file and requires a 3D printer capable of operating at large dimensions – at least 120x200x200 mm of printing capability. Consider a vertical bottom-up position for printing the shell. Support material should be added to the faces of the three tabs. This design file does not present a laser cutting option as it is made of 3D organic shapes not ideal for laser cutting work.</li><li>Batteries, boards, and wheel layers– These design files are composed of three circular platforms that should be placed on the interior of the shell to hold all the electronic components in place. Consider printing the layers horizontally. Support material is needed only on the face of the counter-bore holes of the larger platform. The laser cutting options valid for this design file.</li><li>Jewel layer– This file contains the design that serves to nest the LED jewel that will be attached from the top of the shell. Consider printing the LED nest horizontally with support material. This file does not support a laser cutting option due to its 3D design requirement.</li><li>Glowing fibers layer– This design file contains the plate where the optical fibers should be glued. Consider printing the LED nest horizontally with support material. The laser cutting option is valid for this design file.</li><li>Washer– Washers should be placed between the ‘‘Jewel layer” and the ‘‘glowing fibers layer” to secure this connection. YOLO uses three washers to support this connection, so consider printing 3 parts. The laser cutting option is valid for this design file.</li><li>Motor docking (1) and (2)– Composed of two files that together provide docking for the motors. Support material is not needed for 3D printing. The laser cutting option is valid for ‘‘Motor docking (2)</li></ul><div class="fr-img-space-wrap"><span class="fr-img-caption fr-fic fr-dib" style="width: 788px;"><span class="fr-img-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjE5NjQzOTc2MTgyLTUuanBnIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" style="width: 758px;" class="fr-fic fr-dib"><span class="fr-inner" spellcheck="false">YOLO Parts Lineup</span></span></span></div><p><strong>Assembly</strong></p><p class="fr-img-space-wrap2">The assembly requires the following tools: hacksaw, utility knife, screwdriver set, calipers, scissors, soldering kit(including solder spool, soldering station, wire stripper, diagonal cutters, solder wick for solder removal, soldering vise with a magnifying glass, and a panavise), and glue. 3D print and laser cut the required materials before assembly, configure the voltage transformer with an input of 5.0 V, and stepdown the buck converter output for 1.5 V. Additionally, follow the steps described below</p><p class="fr-img-space-wrap2"><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjE5NjQ0MzY4MTI2LTYuanBnIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" style="width: 723px;" class="fr-fic fr-dib"></p><div><br></div><p>An exploded view of YOLO that supports the understanding of the final robot configuration is present, with close-up views. When the robot is fully assembled, attach a batch of copper tape to the shell of the robot to connect the wire that comes from the capacitive touch sensor. This will enable the robot to respond to touch</p><p><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjE5NjQ0NTI2ODg4LTcuanBnIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" style="width: 788px;" class="fr-fic fr-dib"></p><div><br></div><p>The field of HRI is characterized by multidisciplinarity, for which open access to research tools, such as robots’ hardware and software, is crucial. YOLO presents as a low-purchase and low-maintenance cost robot, that can be used as a tool for research studies with children.&nbsp;</p><p>The open-source hardware of YOLO thus provides opportunities for researchers with and without engineering background to build this robot and further use it to target their own research goals, without depending upon complex robotic platforms.&nbsp;</p><p>To demonstrate how this robot can be applied to academia, researchers can use it as a platform to explore the design of behaviors for a robot aimed at interacting with children. Another example is the usage of this robot by the social and cognitive sciences field as a controllable and programmable tool, to study the developmental aspects of children when interacting with robots.&nbsp;</p><p>Predominantly, the scientific community relies on the usage of off-the-shelf robots as their research platforms when performing studies. Nonetheless, off-the-shelf robotics platforms are generally expensive (with purchase prices ranging from $5:000 to $20:000, or more) and associated with high maintenance costs. In addition, the majority of these robots require special transportation services to be used during field studies, due to their robust size and heavyweight, placing additional costs for academic laboratories. YOLO offers a less expensive yet interesting solution for research.</p><hr><h3><strong>Resources</strong></h3><p>YOLO hardware:&nbsp;</p><p>Alves-Oliveira, P., Arriaga, P., Paiva, A., &amp; Hoffman, G. (2019). Guide to build YOLO, a creativity-stimulating robot for children. <em>HardwareX</em>, <em>6</em>, e00074. Link: <a data-saferedirecturl="https://www.google.com/url?q=https://www.sciencedirect.com/science/article/pii/S2468067218300890&source=gmail&ust=1620392946091000&usg=AFQjCNFJ1CSA2ore98RXkgL20qiBk6G6gw" href="https://www.sciencedirect.com/science/article/pii/S2468067218300890" target="_blank">https://www.<wbr>sciencedirect.com/science/<wbr>article/pii/S2468067218300890</a></p><p>YOLO Software:&nbsp;</p><p>Alves-Oliveira, P., Gomes, S., Chandak, A., Arriaga, P., Hoffman, G., &amp; Paiva, A. (2020). Software architecture for YOLO, a creativity-stimulating robot. <em>SoftwareX</em>, <em>11</em>, 100461. Link: <a data-saferedirecturl="https://www.google.com/url?q=https://www.sciencedirect.com/science/article/pii/S2352711019302468&source=gmail&ust=1620392946091000&usg=AFQjCNEjWO-hCwkOhQw5LtcD72GDY9iFKw" href="https://www.sciencedirect.com/science/article/pii/S2352711019302468" target="_blank">https://www.<wbr>sciencedirect.com/science/<wbr>article/pii/S2352711019302468</a></p>

YOLO is a robot that boosts children’s creativity. It stand for “Your Own Living Robot"

YOLO Robot

<p>This article was discussed in our <a data-saferedirecturl="https://www.google.com/url?q=https://www.wevolver.com/profile/the.next.byte&source=gmail&ust=1640817319915000&usg=AOvVaw2nPRdRwJrXSYYeM9RCFu5x" href="https://www.wevolver.com/profile/the.next.byte" id="isPasted" rel="noopener noreferrer" target="_blank">Next Byte podcast</a>.&nbsp;</p><p><span class="fr-video fr-deletable fr-fvc fr-dvb fr-draggable" contenteditable="false" draggable="true"><iframe height="200px" width="100%" frameborder="no" scrolling="no" seamless="" src="https://player.simplecast.com/ef727514-e071-4310-94c3-ab0d48d9b24d?dark=false" class="fr-draggable"></iframe></span></p><p>The full article will continue below.</p><hr><h1>The Digidog</h1><p>The New York Police department purchased a Spot robot from Boston Dynamics - which the department later named Digidog (personally I feel like this is a missed opportunity to call it RoboPup but oh well) - with the intention of providing an extra layer of safety for the officers called to a scene by having the robot inspect the perimeter.&nbsp;</p><h1>First Day On The Job</h1><p>Digidog's first call to action came when two men were being held captive and tortured in their home by armed intruders who were disguised as plumbers to gain entry. When police were eventually called to the scene by one of the captives who managed to escape, they decided to deploy Digidog and get a better idea of the situation inside the building before officers entered. Unfortunately, the intruders had already escaped but the victims were able to walk away from the situation with non life-threatening injuries.&nbsp;</p><p><br></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/24jufNhuUSI?&amp;wmode=opaque&amp;rel=0" frameborder="0" allowfullscreen="" class="fr-draggable"></iframe></span><br></p><h1>Robots On The Force</h1><p>Although this might be the first utilization of a Boston Dynamics Spot by a police department, authorities have actually been getting extra assistance from robotics for quite some time now. In fact, here are some notable instances in recent years:</p><p>2016 | Dallas, Texas: officers naturalized a target that had murdered 5 officers using a robot</p><p>2015 | San Jose, California: robot delivery of some pizza and a phone helped talk down an armed man off the ledge who was planning on jumping off a bridge</p><p>2014 | Albuquerque, New Mexico: gunman surrendered after robot deployed chemical munitions into the motel room where he had barricaded himself</p><h1>The Opposition</h1><p>Skeptics aren't too eager to embrace the addition of our new metallic friends to the force. Some, including congresswoman Alexandria Ocasio Cortez, have voiced their concerns regarding over-policing of underserved communities and the hefty price tag accompanied by each unit (base price = ~74,000 USD) which could be better spent elsewhere; however, since the Digibot has already started to prove its worth by minimizing endangerment of officers in the field, I suspect it is only a matter of time before we start seeing officers and their robotic partners nationwide.</p>

The New York Police Department has started using a robotic companion to ensure the safety of officers in potentially life-threatening environments

Meet Digidog, The Real Life Robocop

With robots in agriculture becoming more of the norm rather than the exception, the term high-tech farming is no longer an oxymoron. Farming robots are now involved in almost every farming activity, helping farmers fill labor shortages and supermarket shelves simultaneously.

The Future of Agriculture: Adopting More Advanced Robots for Next-Gen Farming

<h2><br></h2><h2>The Business</h2><p>Based in Paintsville, KY, eKAMI is a Haas Technical Education Center (HTEC) that is teaching former coal miners, and others seeking career advancement, CNC machining and other advanced manufacturing skills.</p><h2>Task</h2><p>Training future manufacturing professionals in advanced manufacturing.</p><section><h2>The Challenge</h2><p>Manufacturing faces a significant skilled labor shortage, which is expected to reach 3.5 million unfilled advanced manufacturing job openings by 2025. The industry needs more workers armed with the advanced manufacturing skills. Kathy Walker founded eKAMI, a Haas Technical Education Center, to address this need and also create opportunity for former coal miners who had been left behind by evolving energy policy.</p><p>Over their first few years, eKAMI has trained over a hundred workers on CNC machine programming and operation, as well as other advanced manufacturing skills. Despite a 100% job placement rate, Kathy Walker realized that eKAMI must extend the skills of their graduates to include the ability to design, program, deploy, and manage robotic automation. These skills would multiply the impact that each graduate would have at their future employers, and ensure a bright future for eKAMI graduates.</p><p>However, the robotic landscape is extremely fragmented, with 70+ robot OEMs, each with its own difficult-to-learn programming language. Learning to program basic skills on any robot can take weeks, and require base programming knowledge like C+ or Pascal. This wasn’t a workable solution for eKAMI, as it would require greatly increasing the length of their 16-week training course in order to give their students basic robot programming skills - with just ONE robot brand! eKAMI needed a solution that would allow them to rapidly train their students to program multiple brands of robots, and also learn the related components of automation: automated workcell design, parts presentation, safety, programming and troubleshooting robotic automation, etc.</p><h2>The Solution</h2><p>When the eKAMI team discovered READY Robotics they knew they might have found the solution. READY’s Forge/OS enables operators to program multiple brands of robots from a single platform. With intuitive, flow-chart-based programming, Forge/OS drastically reduces the learning curve, enabling anyone to learn basic programming skills in under a day versus the weeks it takes with each robot’s native programming language. More than that, Forge/OS programming skills transfer to multiple top robotic brands, enabling students to become proficient programming Yaskawa, FANUC and other top brands in just a few days. It was clear to eKAMI leadership that building their automation education around Forge/OS would allow students to quickly expand their skillset to include robot programming. Equally importantly, it would also free up enough time for the students to learn the other skills necessary to successful design and deploy automation. READY was already hard at work on educational materials for their customers, and were happy to build and teach a 3-week automation course for eKAMI.</p></section><section><blockquote>Whether you have some knowledge or zero knowledge with robotics, anyone can be easily taught how to run robots (with Forge/OS).</blockquote></section><blockquote>"Kaylee Maynard" - eKAMI Graduate</blockquote><p><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1606320390823-left-image.jpg" style="width: 788px;" class="fr-fic fr-dib"></p><section><blockquote>Within a couple of hours of robot time I was doing it (programming robots)!</blockquote></section><blockquote>"Mike Cepeda" - eKAMI Graduate</blockquote><section><p><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1606320683797-right-image.jpg" style="width: 788px;" class="fr-fic fr-dib"></p><h2>The Result</h2><p>In August the eKAMI Haas Technical Education Center graduated their first class of students, skilled in both CNC machine programming and operation, as well as automation design, programming and deployment. With a 3-week automation add-on to their 16-week CNC course, this class of eKAMI graduates are able to program machine tools as well as deploy lights out production. Best of all, 3 weeks of automation didn’t just teach them the most basic of robot programming skills that they might have learned through a robot OEM. Instead, they learned how to evaluate potential automation projects, how to design an automated workcell, how to solve parts presentation, how to design and fabricate end-of-arm tooling, how to seamlessly integrate robots with machine tools, how to program robots and deploy automation, how to optimize automated workcells to maximize output, and how to achieve lights out production. With the help of Forge/OS and READY’s automation training curriculum, this class of eKAMI graduates is the first of many that will have the skills needed to help usher in the future of US manufacturing. One where automation is common across manufacturers of all sizes, and one where US manufacturing is once again a global manufacturing leader.</p></section><section><blockquote>If I can do it, anyone can program robots with Forge/OS!</blockquote></section><blockquote>"Tim Miller" - eKAMI Graduate</blockquote><hr><p>Highlights</p><h2>Former coal miners upskilled into automation experts in 3 weeks</h2><p><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1606320602394-1606320602394.png" style="width: 788px;" class="fr-fic fr-dib"></p>

Manufacturing faces a significant skilled labor shortage, which is expected to reach 3.5 million unfilled advanced manufacturing job openings by 2025.

Former coal miners learn to deploy lights out production in 2.5 weeks

We announce the eleven shortlisted finalists of the Make it Real 3D Printing Challenge.

Make it Real 3D Printing Challenge Finalists Announcement

<p dir="ltr">Humanoid robots have been a topic of deep interest amongst engineers due to their ability to perform in environments primarily designed for humans. The field of robotics in general is quite unique as it is made of a fusion of many disciplines from mechanical design to biology. However, the humanoid is particularly fascinating to humankind as it mimics the human form. &nbsp;Generally speaking, a robot can perform a variety of tasks autonomously. Robots can fall into different categories such as industrial, service, social, humanoid, etc..&nbsp;</p><p dir="ltr">Industrial robots, for instance, as the name suggests are designed with the intention of performing a specific task within an industrial context i.e. welding (commonly deployed in automotive plants), commercial vacuum cleaning, and so on. Humanoid robots on the other hand can be merely the design of a particular body part or a complete imitation of the human biomechanics. Through developing humanoids not only we can build machines that are able to operate in human environments but also gain a better understanding of the movements of the human body. This also enables us to design and create more optimal protheses for injured people.&nbsp;</p><h1 dir="ltr">Overall Design Considerations&nbsp;</h1><p dir="ltr">The key consideration associated with the outer appearance of the humanoid robot is how much it will emulate human features. The cornerstone of developing humanoid robots is that they can interact seamlessly with humans. So, understanding how much verbal understanding or response it should be able to perform is paramount. This also determines to what extent these robots can navigate environments natural to humans and operate as such.&nbsp;</p><p dir="ltr">Suffice it to say, when designing a robot, important factors include size, strength and weight. In addition, application plays a crucial role in the development and design process. For example, the infamous <a href="https://www.wevolver.com/wevolver.staff/atlas.robot/" target="_blank">ATLAS robot</a> from Boston Dynamics is designed for physical performances whilst doing motion generation. To perform tasks at a fast pace, humanoid robots should account for stiffness and power output, light weight and more vigilant real-time information processing capabilities. All in all, The design principles can be divided into four following categories:</p><ul><li dir="ltr"><p dir="ltr">Proportions of the body:&nbsp;In order to achieve human-like kinematics and dynamics, similar mass distributions and link lengths are required. Furthermore, humanoids are able to fit better within human environments when accounting for strong body proportion similarities.</p></li><li dir="ltr"><p dir="ltr">Skeletal structure:&nbsp;One important consideration for mimicking the human body is a high degree of fidelity in skeletal structure. Human joints are made up of both single-axis rotation and rolling-sliding joints. This combination allows for sliding and rotation movements between the bones. To emulate flexible upper body movement, spine joints with multiple vertebrae are helpful.</p></li><li dir="ltr"><p dir="ltr">Joint performance:&nbsp;Joint performance determines how the humanoid performs when it comes to the whole-body motion. Two factors impacting joint performance are joint range of motion and power output.</p></li><li dir="ltr"><p dir="ltr">Arrangement of muscles: Muscle arrangement is essential for identifying the muscle-joint-operational mappins, which provide muscle contribution in adequate tendency during whole-body movements. To design a humanoid with human-like muscle arrangement, the humanoid should possess as many muscle actuators as the design space allows for</p></li></ul><p dir="ltr"><span class="fr-video fr-deletable fr-fvc fr-dvb fr-draggable fr-active" contenteditable="false" draggable="true"><iframe src="https://www.youtube.com/embed/_sBBaNYex3E?&amp;wmode=opaque" frameborder="0" allowfullscreen="" class="fr-draggable" style="width: 753px; height: 411px;"></iframe></span><br></p><h2 dir="ltr">Mechanical Structure of Humanoid Robots &nbsp; &nbsp; &nbsp; &nbsp;</h2><p dir="ltr">Human structure has always been the source of inspiration for the humanoid robot's kinematic structure. Thus, most humanoids resemble human features such as head, legs, hands, feet, etc. Moreover, these robots typically include similar Degrees of Freedom (DOFs) to a human being. Evidently, such robots are designed to be human-like. Thus, the center of motion of the robot should be located in the vicinity of its ‘belly button’. The human musculoskeletal system is essentially made up of bones, cartilage, muscles, tendons and skin. It consists of more than 206 bones and 600 muscles.&nbsp;</p><p dir="ltr">Through biomechanics, the musculoskeletal system can be regarded as a mechanical system: the bones can be the rigid links, and the cartilage connects these links. The lower body is generally composed of a waist, upper legs, lower legs and feet in a redundant and human-like kinematic configuration, similar Range of Motion (ROM) and similar joint velocities and accelerations. The ‘legs’ consist of 3 rotation joints with one joint allowing for knee flexion-extension and two other for the flexion-extension and prono-supination of ankles. The type of application determines the grippers used in humanoid robots.&nbsp;</p><p dir="ltr"><span class="fr-img-caption fr-fic fr-dib" style="width: 788px;"><span class="fr-img-wrap"><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1603454172956-robots_0+%281%29.jpg"><span class="fr-inner" spellcheck="false">Image: Boston Dynamics</span></span></span></p><p dir="ltr">Feet are perhaps the most important aspect and they must handle impacts of landing and prevent slippage. Humanoid robots should possess soles with materials and shapes that provide enough friction. For instance, the <a href="https://www.wevolver.com/wevolver.staff/lola/" target="_blank">LOLA robot</a> (a project built with the goal of creating a machine that is capable of stable and fast walking) deploys Sylomer (a special elastomer material) with an elastic modulus that is higher at the heel.&nbsp;</p><p dir="ltr">There’s a wide variety of tasks a humanoid robot can overtake, from extreme or dangerous environments to healthcare. The technology pioneering the progress of these robots continues to grow by leaps and bounds and so does with it their application.</p><h1 dir="ltr">Sources &amp; Further Reading&nbsp;</h1><p dir="ltr"><a href="https://www.thelancet.com/journals/ebiom/article/PIIS2352-3964(19)30534-1/fulltext" target="_blank">Human–robotic interfaces to shape the future of prosthetics</a></p><p dir="ltr"><a href="http://ai.stanford.edu/~rxzhang/Mechanical_Design_And_Control_System.pdf" target="_blank">Mechanical Design And Control System Configuration of a Humanoid Robot</a></p><p dir="ltr"><a href="https://www.sciencedirect.com/science/article/pii/S0928425709000400" target="_blank">Humanoid robot Lola: Design and walking control</a></p>

Chair of Applied Mechanics, Technical University of Munich (TUM)

Humanoid robots require deeper consideration than just mimicking humans.

Design Considerations for Humanoid Robots

<p dir="ltr">After the introduction of digital tools in design and creative disciplines almost three decades ago, there has been an ongoing interest in connecting the physical and the digital worlds. Most of these hybridizations have been directly influenced by the industrial/engineering origins of some of these digital technologies. For instance, using CNC/Robotic arms as tools for fabricating digitally designed artifacts or seeking time/cost-efficient production methods are a few examples.&nbsp;</p><p dir="ltr">Although there is a tremendous value in this "tool-oriented" perspective, there seems to be a missed opportunity to use these digital/physical platforms, not only as "tools"--for translating one platform to another, but to combine them into a synthesized mixed design/experience "medium"; a mixed medium where it benefits from both digital and physical to curate a new design process, artifact or experience.&nbsp;</p><p dir="ltr">In one of their latest experimental research projects Mixed Robotic Interface (MRI Γ), researchers at the Robotically Augmented Design (RAD) Lab at the College of Architecture and Environmental Design (CAED) at Kent State University, explored potentials of robotics and augmented reality technologies as possible mediums for hybrid cyberphysical storytelling and atmosphere making. Using the natural power of robotics and AR as platforms that co-exist in both digital and physical environments, MRI Γ explores possibilities of creating/curating a hybridized experience arrangement, to narrate a story with actual and virtual actors and setups.&nbsp;</p><p dir="ltr">Developed as part of the MRI Γ research investigation led by Ebrahim Poustinchi assistant professor of architecture and the founder/director of the RAD Lab, graduate students Bradly Bowman and Trever Swanson formed a performance setup for storytelling, using a combination of physical images/posters, AR content as the performance set/stage, and an industrial robotic arm as the performer.&nbsp;</p><p dir="ltr"><span class="fr-video fr-deletable fr-fvc fr-dvb fr-draggable fr-active" contenteditable="false" draggable="true"><iframe src="https://player.vimeo.com/video/447412833" frameborder="0" allowfullscreen="" class="fr-draggable" style="width: 778px; height: 565px;"></iframe></span><br></p><p dir="ltr">The story—called “Child Robot!,” is based on a para-fictional scenario speculating on a future child room where the robot arm plays the role of the child, illustrating a futuristic/post-human ambiance. As a cyberphysical storytelling arrangement that simultaneously uses element from the actual and virtual to narrate, precision, and coordination between the digital and the physical is crucial. The setup is a product of a precise calibration of the physical robot arm—KR6 R900 Sixx KUKA robot, the physical and virtual room, and the virtually animated toys in the scene (child-robot room) coordinated in a responsive time-based manner. &nbsp;Using augmented reality procedures in conjunction with the physical setup, “Child Robot!” enables the users to experience a blend of reality, fantasy, actual and virtual through a custom-made software application, curated motion, and a physical scene.&nbsp;</p><p dir="ltr">Powered using Unity 3D game engine, custom animation based robotic motion-control platform—Oriole, and image tracking techniques, “Child Robot!” facilitates its audience to experience the story—and its hybridized digital/physical content, from multiple points of view and perspectives. Liberating the experience from both completely immersive digital experience—VR experience, and a physical single perspective performance—regular physical experience, MRI blurs the boundaries between the audience, the performer, and the performance setup.&nbsp;</p><p dir="ltr">To synchronize and coordinate the robot's movement and the digital game events in the AR platform, two sets of data have been used. A physical image-tracker works as an anchor for the digital—AR, content, and establishes a coordinate system for the scene's virtual overlay. This tracker enables the coordination between the robot's static position and the static framing of the digital setup. However, for the robot—from the physical world, to interact with the digital game objects, there is a need for realtime communication. Throughout the experience/performance, the robot continually sends its realtime position/orientation (TCP Plane data) to Unity 3D game engine. This information enables interaction between the game objects—toys, and the robot.</p><p dir="ltr">To overlay the digital renderings—AR material, over the real/physical footage of the setup without overwriting the robot itself, RAD Lab researchers used greenscreen techniques to separate the robot from its physical context. Later in the process by using a custom-made “Child Robot!” AR application, and employing Chroma Key image processing techniques, the green background has been replaced with a coordinated digital content in realtime.&nbsp;</p><p dir="ltr">While holding a device—e.g., cellphone, equipped with the custom-made “Child Robot!” AR application, the audience can walk through the MRI Γ setup, physically interact with the robot—the performer, and watch the playful interaction of the physical robot and the digital room.&nbsp;</p><p dir="ltr"><span class="fr-img-caption fr-fic fr-dib" style="width: 788px;"><span class="fr-img-wrap"><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1597308421054-003.jpg"><span class="fr-inner" spellcheck="false">The medium for experiencing the “Child-Robot! ”is a heterogeneous mixture of reality and its augmentation through the lens of a custom-made cell-phone app. The robot-arm as the performer/narrator of the story remains connected to the world of physicality while the digital window augments its surroundings.</span></span></span></p><p>Although MRI Γ research—and “Child Robot!” as one of its examples, are at the early stages of experimentation and development, one of the main aims of this research is to move beyond immersion—both complete digital immersion (VR) or physical. It seeks a hybrid medium—a mixed interface, where the physicality of the sound, presence, and breeze of the robot’s motion, combines with the digitality of rendered animated content of AR to narrate one consistent story and create/curate one cohesive cyberphysical experience.</p><p><span class="fr-img-caption fr-fic fr-dib" style="width: 788px;"><span class="fr-img-wrap"><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1597308435023-002.jpg"><span class="fr-inner" spellcheck="false">Robot arm interacting with virtual objects through synchronized digital/physical realtime animation.</span></span></span></p>

Using Chroma Key image processing techniques in Unity 3D, the green screen is being replaced by the animated digital scene.

Using AR and robotics as Storytelling Mediums

Mixed Robotic Interface (MRI Gamma)

<p>If you are reading this, you've perhaps come across these discussions on the internet: "Is AI going to take away your jobs?", "Will AI replace humans by becoming better than them in every task?", and also some exaggerated, melodramatic versions like: "AI armageddon!!" and "Evil Robots will take over!", etc. To add fuel to fire, Hollywood paints a vivacious picture of just that. I haven't yet come across a robot/AI movie that ends well, have you? All this said and done, have you ever wondered where we are in that path? Is everything Hollywood saying true? Will I indeed lose my job? Will robots take over the world? Whoaaa... It's ok, take a breather! You're going to be fine! The world will be fine!</p><p>In this article, I will give you my take on where we are possibly headed and what could be a plausible reality we can prepare for. We will also look at the contribution of THINC Lab, where the best minds come together to create AI solutions to enrich and enhance human life. But first, let's just pause and mull over the remarkable progress mankind has made since we stepped on the planet... From not knowing how to light a fire, we've created rockets that blast off into space using fire, from not knowing how to verbally communicate with each other, we've sent out voyager satellites hoping to communicate with other intelligent beings out there! and throughout this incredible journey, mankind has gone through a phenomenal evolution, but the one thing we've always been very good at is survival! Everything we do, we do to survive and grow, emotionally, socially, culturally, and ofcourse, intellectually. So why are we creating something that might threaten our existence one day? The short answer, WE'RE NOT! At least that's not the intent!&nbsp;</p><p>I personally feel there's a lot of miscommunication or just a lack of communication between the actual architects who work on building these systems and the stakeholders/consumers who are on the receiving end of it. You might've heard of people referring to AI as a double-edged sword, but aren't we humans building the swords in the first place? Couldn't we alter its design with time as needed? Absolutely. Everytime a new scientific invention is made, people fear the loss of jobs and a detrimental shift in paradigm. It doesn't have to be that way as long as the Engineers and Scientists building it are cognizant of the journey we are on. But then people point me to the industrial revolution that caused a massive loss of jobs and turned people's lives topsy turvy, I say look at the world now, people adapted, developed and switched to better domains that demanded the innate humanness. Change is always messy and it takes time for the chaos to reach a steady-state, but that fear of change can't hamper our curiosity and imagination, for like Einstein said, "Knowledge is limited, but imagination encircles the world".</p><p>But hey, I still haven't told you where we are on this journey. Although there's some debate about it, the common consensus is that AI began sometime in the early 1950s. From the time of Alan Turning (also called the Father of AI, and there is an award attributed to his name - you guessed it, the Turing award), we have made some significant strides in the field of AI and with Quantum Computers coming up, the power of AI seems limitless. We live in the age of Neural Networks, where a bunch of neurons (yes, that is an analogy of the human brain adapted in the NN architecture) crunch data using the fastest computers and learn "patterns". Using these "patterns" or optimized objective function model(s), they can make predictions about the future or recognize other objects that belong to the class of data they were trained on. Some eerily advanced applications include <a href="https://en.wikipedia.org/wiki/Deepfake" rel="noopener noreferrer" target="_blank">Deep Fakes</a>, <a href="https://machinelearningmastery.com/impressive-applications-of-generative-adversarial-networks/" rel="noopener noreferrer" target="_blank">Generative Networks</a>, etc. Also in the field of Robotics, we've handsomely advanced to create robots that can <a href="https://blog.aboutamazon.com/transportation/whats-next-for-amazon-scout" rel="noopener noreferrer" target="_blank">deliver packages</a> to your house, <a href="https://www.tesla.com/models" rel="noopener noreferrer" target="_blank">cars</a> that can drive themselves and we're also in the process of building <a href="https://www.bostondynamics.com/" rel="noopener noreferrer" target="_blank">military robots</a>. So, if AI and robots can do all this, what promise do we have that they will not outperform humans and take over jobs? Honestly, we don't. It would be foolish to think that some jobs would not be displaced/restructured by AI and Robotics.&nbsp;</p><p>But since we can see it coming now, we can prepare for it, adapt, and make a smooth transition into this new era. But for the most part, the intent of AI/Robotics is to work alongside humans to aid and improve human life. One such example is the project our researchers at THINC Lab are involved in. At THINC, we strive to create cutting-edge AI solutions that can be seamlessly integrated into industries and workplaces. A real-life example of such a solution is the one we are currently developing with the funding provided by the National Science Foundation to create intelligent robots that work alongside humans in an onion sorting plant to sort onions quickly and efficiently.&nbsp;</p><p><span class="fr-img-caption fr-fic fr-dib" style="width: 375px;"><span class="fr-img-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiAid2V2b2x2ZXItcHJvamVjdC1pbWFnZXMiLCAia2V5IjogImZyb2FsYS8xNTk1MTc1MjIwNDUwLUlNR18yMDE5MDQyMl8xNzU0NTEwNjIuanBnIiwgImVkaXRzIjogeyJyZXNpemUiOiB7IndpZHRoIjogOTUwLCAiZml0IjogImNvdmVyIn19fQ=="><span class="fr-inner">Director of THINC Lab Dr.Doshi examining the robot during setup.</span></span></span><span class="fr-img-caption fr-fic fr-dib" style="width: 378px;"><span class="fr-img-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiAid2V2b2x2ZXItcHJvamVjdC1pbWFnZXMiLCAia2V5IjogImZyb2FsYS8xNTk1MTc1NjQyOTA2LXNhd3llci5KUEciLCAiZWRpdHMiOiB7InJlc2l6ZSI6IHsid2lkdGgiOiA5NTAsICJmaXQiOiAiY292ZXIifX19"><span class="fr-inner">What the sorting looks like within a computer simulation.</span></span></span><span class="fr-img-caption fr-fic fr-dib" style="width: 383px;"><span class="fr-img-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiAid2V2b2x2ZXItcHJvamVjdC1pbWFnZXMiLCAia2V5IjogImZyb2FsYS8xNTk1MTc1ODkwNzM4LVNjcmVlbnNob3QgZnJvbSAyMDIwLTA1LTI4IDE4LTM1LTMyLnBuZyIsICJlZGl0cyI6IHsicmVzaXplIjogeyJ3aWR0aCI6IDk1MCwgImZpdCI6ICJjb3ZlciJ9fX0="><span class="fr-inner">Image showing how the robot perceives the world around it.<br></span></span></span></p><p>You may say, "Hey, how hard can that be?", but think about this, humans possess generations of evolutionary intelligence combined with a unique set of sensory apparatus that help us feel, see and understand if an onion is good/bad for consumption. However, all a Robot sees is a mathematical representation of the world around it, and teaching it to do this task involves significant challenges, especially when you expect it to collaborate safely with humans. We are currently developing a Neural Network - <a href="https://arxiv.org/abs/1905.04380" rel="noopener noreferrer" target="_blank">SA-Net</a> for seeing and interpreting the visual information and combining it with our unique in-house recipe of a classical AI algorithm that would tell the robot what to do and how to do it. If you are interested to read more about what we do, check out our <a href="http://thinc.cs.uga.edu" rel="noopener noreferrer" target="_blank">website</a>. Thanks for taking this trip on my train of thought!</p>

And how are AI scientists involved in this?

How far along are we in the AI-Human "race"?

<p>What if running the 26.2 miles of a marathon only felt like running 24.9 miles, or if you could improve your average running pace from 9:14 minutes/mile to 8:49 minutes/mile without weeks of training? Researchers at the<a href="http://seas.harvard.edu/" target="_blank">&nbsp;Harvard John A. Paulson School of Engineering and Applied Sciences&nbsp;</a>(SEAS) and the&nbsp;<a href="http://wyss.harvard.edu/" target="_blank">Wyss Institute for Biologically Inspired Engineering at Harvard University</a> have demonstrated that a tethered soft exosuit can reduce the metabolic cost of running on a treadmill by 5.4 percent compared to not wearing the exosuit, bringing those dreams of high performance closer to reality.</p><p>&quot;<em>Homo sapiens</em> has evolved to become very good at distance running, but our results show that further improvements to this already extremely efficient system are possible,&rdquo; said author Philippe Malcolm, former postdoctoral research fellow at SEAS and the Wyss Institute, and now Assistant Professor at the University of Nebraska, Omaha, where he continues to collaborate on this work.</p><p>The study appears today in&nbsp;<a href="http://robotics.sciencemag.org/content/2/6/eaan6708" target="_blank"><em>Science Robotics</em></a>.</p><p>Running is a naturally more costly form of movement than walking, so any attempt to reduce its strain on the body must impose a minimal additional burden. The soft exosuit technology developed in the lab of <a href="https://www.seas.harvard.edu/directory/walsh" target="_blank">Conor Walsh</a>, John L. Loeb Associate Professor of Engineering and Applied Sciences at SEAS and Wyss Core Faculty member, represents an ideal platform for assisted running, as its textile-based design is lightweight and moves with the body.</p><p><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1577273403478-Exosuit-Running.jpg" style="width: 1114px;" class="fr-fic fr-dib"></p><p>A team of scientists in Walsh&rsquo;s lab led by postdoctoral fellow Giuk Lee performed the study with an exosuit that incorporated flexible wires connecting apparel anchored to the back of the thigh and waist belt to an external actuation unit. As subjects ran on a treadmill wearing the exosuit, the actuation unit pulled on the wires, which acted as a second pair of hip extensor muscles applying force to the legs with each stride. The metabolic cost was measured by analyzing the subjects&rsquo; oxygen consumption and carbon dioxide production while running.</p><p>The team tested two different &ldquo;assistance profiles,&rdquo; or patterns of wire-pulling. One was based on human biology that applied force starting at the point of maximum hip extension observed in normal running. The other was based on a simulation of exoskeleton-assisted running from<a href="http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0163417#abstract0" target="_blank">&nbsp;a group at Stanford University</a>, which applied force slightly later in the running stride. The Stanford study suggested that the optimal point to provide assistive force might not be the same as the biological norm. Confirming this suspicion, Lee and colleagues found that the simulation-based profile outperformed the biology-based profile in reducing metabolic cost by about a factor of two.</p><p>As the authors note in the paper, &ldquo;Our finding supports a paradigm shift toward the concept that mimicking our current understanding of biology is not necessarily always optimal.&rdquo;</p><p>To clarify why applying force later in running stride improved metabolic expenditure so dramatically, the team analyzed what was happening to the subjects&rsquo; other joints when their hip joints were being assisted by the actuating wires. They found that the simulation-based profile also affected knee extension and the forces between the foot and the ground, while the biology-based profile did not.</p><p>&ldquo;The biological profile only takes into account the amount of torque in the hip joint, but the human body is not a series of independently acting parts - it&rsquo;s full of muscles that act on multiple joints to coordinate movement,&rdquo; said Lee. &ldquo;Applying force to the hip affects the whole body system, and we need to consider that in order to give the best assistance.&rdquo;</p><p>While the study&rsquo;s results suggest that the increased biomechanical involvement of the simulation-based assistance profile is responsible for the reduction in metabolic cost, further biomechanical analyses are needed to confirm its findings.</p><p>The team hopes to continue this research to reduce the metabolic cost of running even more.</p><p>&ldquo;We only tested two actuation profiles in this study, so it will be interesting to see how much more the cost of running can be reduced with further optimization of the system,&rdquo; said Malcolm.</p><p>&ldquo;Our goal is to develop a portable system with a high power-to-weight ratio so that the benefit of using the suit greatly offsets the cost of wearing it. We believe this technology could augment the performance of recreational athletes and/or help with recovery after injury,&rdquo; said Lee.</p><p>The days of a battery-powered exosuit for high-performance runners are still beyond the horizon, as the actuator unit (including motors, electronics, and power supply) in this study was off-board, but the authors say technology is moving toward making an untethered assistive exosuit possible in the near future.</p><p>&ldquo;This study is a great example of how our team can quickly leverage its experience and expertise in wearable technology and biomechanics to demonstrate how human performance can be improved in new application areas,&rdquo; said Walsh.</p>

Researchers have demonstrated a tethered soft exosuit that can reduce the metabolic cost of running on a treadmill by 5.4 percent compared to not wearing the exosuit (Image courtesy of The Wyss Institute)

New robotic exosuit could push the limits of human performance and lead to new wearable technologies for athletes and consumers

Tethered soft exosuit reduces the metabolic cost of running

<p><strong>One key element of the future production landscape is highly flexible Human-Robot Collaboration (HRC). This article describes 3 case studies which illustrate how collaborative robots change the way people will manufacture.<br><br>KUKA, one of the world&#39;s largest suppliers of industrial robots, has written a free, downloadable white paper called &quot;</strong><a href="http://bit.ly/kuka-whitepaper-US" rel="noopener noreferrer" target="_blank"><strong>Shaping future production landscapes</strong></a><strong>.&quot; The white paper describes &#39;megatrends&#39; that impact manufacturing and 3 key factors to designing production infrastructures that can meet the requirements of the future.&nbsp;</strong></p><hr><h1>Intro</h1><h2>Human-robot collaboration&nbsp;</h2><p>In human-robot collaboration, the robot assists the human operator. This means: The machine does not replace the human, but complements his capabilities and relieves him of arduous tasks. These can include overhead work, for example, or the lifting of heavy loads. Autonomous, collaborative robots are also used to supply production workstations.</p><p>In the ideal &#39;factory of the future&#39; there is no separation between automated and manual workstations. Humans and robots collaborate optimally &ndash; without separation and without safety fencing.</p><h2>Human-robot collaboration: the advantages</h2><p>HRC-capable robots (often labeled &#39;Collaborative Robots&#39; or &#39;Cobots&#39;) are rendered mobile and capable of performing different tasks flexibly using mobile platforms. They can thus be individually deployed at whatever location and for whatever purpose corresponds to production requirements &ndash; dependent on the required batch size, for example.<br>Furthermore, these types of robots offer maximum flexibility by means of so-called spontaneous automation: They support a user as an assistant in the case of workload peaks and resource bottlenecks in production operations.</p><p>Human-robot collaboration changes industrial production and manufacturing and can bring a number of advantages over traditional automation or manual labor:</p><ul><li>Increased flexibility in production</li><li>Relief of employees by performing ergonomically unfavorable work steps that could not previously be automated</li><li>Reduced risk of injuries and infections</li><li>High-quality performance of reproducible processes &ndash; without requiring type-specific or component-relevant investment.</li><li>Increased productivity and improved system complexity thanks to integrated sensors</li></ul><p><br></p><h1><strong>HRC Case Studies</strong></h1><p>To give you an idea of the impact of collaborative robots in the workspace we&#39;ve gathered 3 case studies to show you the impact of these robots and the challenges they overcome.</p><p><br></p><h2><strong>Case Study 1: Lightweight robots supporting operators in adhesive bonding processes.</strong></h2><h2><em><strong><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1566292124494-KUKA+LBR+iiwa+Duerr.jpg" class="fr-fic fr-dib" style="width: 766px;"></strong></em></h2><p><em>All images courtesy KUKA</em></p><p>System and machine builder <a href="https://www.durr.com/en/" rel="noopener noreferrer" target="_blank">D&uuml;rr</a> uses two sensitive LBR iiwa robots in final assembly. Without a safety fence, they perform key tasks in the application of adhesive beads. This improves quality, saves time and lowers unit costs.</p><div data-ignore="true">At the company, humans and robots work together in final assembly without the need for physical safeguards. Sensitive lightweight robots are used for the adhesive bonding of GPS antenna covers and tanks in the vehicle body. The solutions improve the quality of the adhesive bonding result, save time and lower unit costs.&nbsp;</div><p><strong>The LBR iiwa detects any fault that arises when the GPS antenna cover is loaded</strong></p><p>For application of adhesive to the GPS antenna cover, the assembly worker places the workpiece into the robot gripper manually. The gripper sucks onto it and moves it to the adhesive nozzle on the application tower. The robot slowly moves the GPS antenna cover upwards to the application nozzle. If it encounters an obstacle, it moves back a little thanks to its collision detection ability and starts the motion over from the beginning. Only after three attempts does it move back into the starting position. Otherwise, the adhesive application process starts and the adhesive bead is applied carefully while the robot executes the path. Finally, back at the starting point, the assembly worker removes the GPS antenna cover and mounts it on the vehicle. &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;&nbsp;</p><p><span class="fr-video fr-deletable fr-fvc fr-dvb fr-draggable" contenteditable="false" draggable="true"><iframe src="https://www.youtube.com/embed/PdAXCNtXabc?&wmode=opaque" allowfullscreen="" class="fr-draggable" style="width: 767px; height: 429px;" frameborder="0"></iframe></span></p><p><strong>Precise adhesive beads thanks to the seventh axis of the LBR iiwa</strong></p><p>For the automated process of gluing the tank into place in final assembly, the skilled worker guides the tank to a turntable with the aid of a manipulator. There he cleans it, moves it into the correct position and transfers it directly to the robot for further processing. When the starting point is reached on the stationary tank, the nozzle opens under pressure so that no air bubbles are created. The robot then applies the adhesive bead onto the tank very evenly. Sensors on the application head monitor the accurate height of the bead. Thanks to its seventh rotating axis, the robot glues in a radius of 360 degrees &ndash; without re-orienting or having to interrupt the bead. As soon as the robot has completed its task, the skilled worker fits the tank into the vehicle body at the specified position.</p><p><a href="http://www.durr.com/" target="_blank">D&uuml;rr</a> has already sold ten systems in the field of adhesive bonding, and is working on HRC solutions for the robot-based adhesive bonding of small washers and other components.</p><p><br></p><h2><strong>Case Study 2: Human-robot collaboration during the adjustment of headlights.</strong></h2><p><strong><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1566292345359-Fahrassistenzsystemprfstnden_Drr.jpg" style="width: 766px;" class="fr-fic fr-dib"></strong></p><p>To reduce the strain on the workers, two KUKA LBR iiwa cobots perform the unergonomic task of adjusting fog lights at Ford. Parallel to this, human colleagues can adjust the main headlights.</p><article data-ignore="true"><p><strong>The initial situation</strong></p>In the automotive industry, vehicle fog lights have previously been adjusted manually by human operators. While stooped over, the human operator searches for the hard-to-reach opening for the adjusting screws in the bumper area. Once he has found the adjusting screw, he positions the tool on it extremely carefully without damaging the adjustment opening on the bumper.</article><article data-ignore="true"><p><strong>The solution</strong></p>To provide relief for the human worker during this ergonomically unfavorable work and simultaneously achieve improved of the adjustment, Ford VOME (Vehicle Operation Manufacturing Engineering), D&uuml;rr Assembly Products and KUKA have jointly developed an innovative alternative concept for the conventional process as part of a feasibility study. The solution called for the use of the KUKA LBR iiwa cobot. The sensitive robot adjusts the fog lights fully automatically. In doing so, humans and robots work on the same vehicle without additional safety equipment.</article><p><span class="fr-video fr-deletable fr-fvc fr-dvb fr-draggable" contenteditable="false" draggable="true"><iframe src="https://www.youtube.com/embed/9HWY-CLR3Co?&wmode=opaque" allowfullscreen="" class="fr-draggable" style="width: 766px; height: 419px;" frameborder="0"></iframe></span></p><p><strong>Teamwork</strong></p><p>While the human operator adjusts the conventional headlights, the cobot adjusts the hard-to-reach fog lights. Two lightweight robots are already in use in four headlight/driver assistance system test stands for the new Ford Focus C519 at the Ford plant in Saarlouis. &nbsp; &nbsp; &nbsp; &nbsp;</p><p><strong>The secret of the HRC concept</strong></p><p>With its size and joint torque sensors, the robot detects contact immediately and reduces its level of force and speed instantly. These characteristics make it possible to work with human operators in confined spaces. Furthermore, with its high-performance servo control, the cobot detects contours quickly under force control. In combination with an end effector specially developed by D&uuml;rr (consisting of a camera system and an adjustment screwdriver), the robot positions the adjustment screwdriver according to the hole coordinates acquired by the camera system. The adjusting screw itself is located approximately 80 mm behind the hole in the bumper. Thanks to sensitive movements, the cobot optimally positions the adjustment tool into the screw head.</p><p><strong>Redundant safety</strong> is provided by the HRC-capable, heavy-duty aluminum gantry from D&uuml;rr Assembly Products &ndash; on which the KUKA LBR iiwa is positioned in a ride-on installation above the light collecting box directly in front of the headlight to be adjusted. The advantage here: additional robot positioning can be dispensed with.</p><p><br></p><h2><strong>Case Study 3: Using a cobot to install pump wells. <br></strong></h2><h2><strong><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1566292578317-KUKAflexFellow_Platzierung_im_Arbeitsbereich_IMG_3628.jpg" style="width: 766px;" class="fr-fic fr-dib"></strong></h2><p>In the future, BSH Hausger&auml;te GmbH will be using a <a href="https://www.kuka.com/en-us/products/mobility/mobile-robot-systems/kuka-flexfellow" rel="noopener noreferrer" target="_blank">KUKA flexFELLOW</a> to carry out the installation of pump wells on the dishwasher production line.&nbsp;</p><article data-ignore="true"><p><strong>The initial situation</strong></p><p>The <a href="https://www.bsh-group.com/us/" rel="noopener noreferrer" target="_blank">BSH Hausger&auml;te</a> GmbH factory in Dillingen, Germany is Europe&rsquo;s largest and most modern factory for dishwashers. Around 10,000 machines leave the production line every day in two-shift operation. There are a total of seven production lines for manufacturing dishwashers.</p><p>Screw fastening of the pump covers in the dishwasher is an ergonomically inconvenient task as the employee has to lean into the dishwasher housing to perform the task. The mechanical assistant now takes on this monotonous step in the process.&nbsp;</p><p><strong>The task &nbsp; &nbsp;&nbsp;</strong>&nbsp; &nbsp;</p><p>Its contribution has already been tested successfully over several months as part of a pilot system. In addition to relieving the human operator of a cumbersome task, the robot also contributes by documenting its work and indicating whether the screws have been tightened correctly or not.</p><p><strong>The solution &nbsp; &nbsp; &nbsp; &nbsp;</strong></p><p>The key challenge when designing the system was to ensure that the manual workstation remained unchanged and that the automation solution could be integrated into the existing assembly line. Its sensitive and responsive capabilities enable it to recognize if, for example, a pump well component is not positioned correctly and then presses it into place.&nbsp;</p><span class="fr-video fr-deletable fr-fvc fr-dvb fr-draggable" contenteditable="false" draggable="true"><iframe src="https://www.youtube.com/embed/jtkJQJXO7AQ?&wmode=opaque" allowfullscreen="" class="fr-draggable" style="width: 767px; height: 413px;" frameborder="0"></iframe></span><p>The advantages of using a cobot:</p><ul><li>Reduction of the manual workload in an ergonomically inconvenient task</li><li>Automated documentation of work steps</li><li>Indication of correct screw fastening operation</li></ul><p>Using its responsive capabilities, the cobot can then calibrate itself independently at its workstation, use a search run mode to find the screw positions, re-apply pressure to the component if it is not correctly positioned and finally tighten the four screws.</p><p><br></p><hr><h1><br></h1><h1>Further Reading</h1><ul><li>Specifications of the <a href="https://www.wevolver.com/wevolver.staff/kuka.lbr.iiwa/" rel="noopener noreferrer" target="_blank">KUKA LBR iiwa</a></li><li>&quot;<a href="http://bit.ly/kuka-future" rel="noopener noreferrer" target="_blank"></a><a href="http://bit.ly/kuka-whitepaper-US" rel="noopener noreferrer" target="_blank">Shaping future production landscapes,</a>&quot; KUKA&#39;s white paper describing the &#39;megatrends&#39; that impact manufacturing and 3 key factors (<strong>people, technology and adaptation</strong>) to designing production infrastructures that can meet the requirements of the future.</li></ul><p><br></p><p><br></p></article>

In human-robot collaboration, human and machine work hand in hand. The human operator controls and monitors production, the robots perform the physically strenuous work. Both contribute their specific capabilities.

Human-robot collaboration: 3 Case Studies

The android called Kannon Mindar. Photo by Richard Atrero de Guzman 