Video Friday this is your weekly selection of amazing robotics videos collected by your friends on IEEE spectrum robotics. We're also posting a weekly calendar of upcoming robotics events for the next few months. Please send us your events to turn on.
IKRA 2026: 1–5 June 2026, VIENNA
Enjoy today's videos!
Unlike existing hybrid designs, Duawlfin eliminates the need for additional drives or propeller ground propulsion system, using only standard quadcopter motors and the introduction of differential transmission with one-way bearings. The seamless transition between air and ground modes further highlights the practicality and efficiency of our approach for applications such as urban logistics and internal navigation.
[ HiPeR Lab ]
I appreciate the softness of the NEO's design, but these fingers feel terribly fragile.
[ 1X ]
Imagine that you are reaching into your backpack looking for your keys. Your eyes guide your hand toward the opening, but once inside, you rely almost entirely on your sense of touch to distinguish between your keys and your wallet, phone, and other items. This seamless transition between sensory modalities (knowing when to rely on vision rather than touch) is something that humans do without much effort, but that robots struggle with. The problem is not just the presence of multiple sensors. Modern robots are equipped cameras, tactile sensorsdepth sensors and much more. The real challenge is **how to integrate these different sensory streams**, especially when some sensors provide scant but important information at key moments. Our solution is based on rethinking how we integrate modalities. Instead of forcing all sensors into a single network, we train separate expert policies for each modality and learn to combine their policy-level action predictions.
Inter-university cooperation is represented through [ GitHub ]
Thank you, Haonan!
Glad (a little late) Halloween from the pollen of robotics!
[ Pollen Robotics ]
In collaboration with our colleagues at Iowa State and the University of Georgia, we have implemented pipe scanning technology. worm robot for field testing. Watch it crawl through corrugated drainage pipes in a creek and across a smooth section of an underground drainage system.
[ Paper ] from [ Smart Microsystems Laboratory, Michigan State University ]
Heterogeneous teams of robots working in realistic environments often have to perform complex tasks that require collaboration and adaptation to information received online. Because robot teams often work in an unstructured environment—an uncertain open world without prior maps—subtasks must be based on the robots' capabilities and the physical world. We present SPINE-HT, a framework that addresses these limitations by grounding LLM reasoning abilities in context heterogeneous team of robots through a three-step process. In real experiments with the jackal of the clear path, the husky of the clear path, Boston Dynamics Spot and high altitude UAVOur method achieves an 87% success rate in missions that require analyzing the robot's capabilities and refining subtasks using online feedback.
[ SPINE-HT ] from [ GRASP Lab, University of Pennsylvania ]
[ Astribot ]
In two articles published in Make a difference and Advanced Science, a group of scientists from the department of physical intelligence Max Planck Institute Intelligent systems in Stuttgart, Germanydeveloped control strategies to influence the movement of self-propelled oil droplets. These oil droplets mimic single-celled microorganisms and can autonomously solve a complex maze by following chemical gradients. However, it is very difficult to integrate external disturbances and use these droplets in robotics. To solve these problems, the team developed magnetic droplets that still have life-like properties and can be controlled by external factors. magnetic fields. In their work, the researchers showed that they are capable of controlling the movement of droplets and using them in micro-robotics applications such as freight transport. transport.
Everyone has dreamed of a personified avatar! Whole body teleoperation and the whole body data collection the platform is waiting for you to try it out!
[ Unitree ]
This is not humanoidbut for now it does useful things safely and probably doesn't cost much to buy or operate.
[ Naver Labs ]
This article presents a curriculum based curriculum. reinforcement learning basis for training accurate and highly efficient jumping rules for the Olympus robot. Separate policies are developed for vertical and horizontal jumps using a simple but effective strategy. Experimental validation demonstrates horizontal jumps up to 1.25 m with centimeter accuracy and vertical jumps up to 1.0 m. In addition, we show that with minor modifications the proposed method can be used for training. omnidirectional jumping.
[ Paper ] from [ Autonomous Robots Lab, Norwegian University of Science and Technology ]
Heavy payloads are no problem for him: the new KR TITANIUM The Ultra moves a payload of up to 1500kg, making heavy lifting extreme in CRY briefcase.
[ Kuka ]
Good luck getting all the sand out of this robot. Perhaps a good oil bath would do?
[ DEEP Robotics ]
This CMU RI seminar is led by Yuke Zhu from the University of Texas at Austin on the topic “Towards a Universalist.” Humanoid robots: Recent advances, opportunities and challenges.”
In the era of rapid progress of artificial intelligence, the use of accelerated computing and big data opened up new opportunities for the development of general specialists artificial intelligence models. Since artificial intelligence systems such as ChatGPT Having demonstrated remarkable achievements in the digital sphere, we are forced to ask: can we achieve similar breakthroughs in the physical world – creating versatile humanoid robots capable of performing everyday tasks? In this talk, I will outline our principles and data-driven research approaches to enable general-purpose robot autonomy in the open world. I will present our recent work on using real, synthetic and web data for training foundation models for humanoid robots. In addition, I will discuss the opportunities and challenges of creating the next generation of intelligent robots.
[ Carnegie Mellon University Robotics Institute ]
Articles from your site
Related articles on the Internet
