Abdu Khair

Last-mile deliveries in urban areas are causing an increase in traffic problems and supply chain issues. To address this, researchers at ETH Zürich have developed a robotic delivery system that uses a hybrid wheeled-legged locomotion to improve efficiency and battery life. With this system, the robot can move from the street to alternate pathways and smoothly transition between walking and driving modes. The robot has been trained using reinforcement learning techniques to navigate over obstacles like stairs and pedestrians. It has been tested for large-scale navigation, covering over 10 kilometers in Zurich, Switzerland, and Seville, Spain. The hybrid wheeled-legged platform has successfully addressed many challenges associated with robotic urban mobility. Read the research here: https://lnkd.in/ekVBEf6Y

29

🤖 Mobile Robot localization vs Laser Scanners 📡 📄 Technology Friday #23 Laser scanners are widely used devices in a huge number of industries and applications/products 💡 🤖 In mobile robotics, like Mobile Industrial Robots A/S - laser scanners are placed in the corners of the robot and perform two, key functions 🗝 : 1️⃣ Take care of safety by detecting obstacles effectively stopping the robot in critical situations - the range of reaction to an obstacle increases dynamically as the speed of the robot increases ❗ 2️⃣ They are the "eyes" of the robot - readings from the scanners form a picture of what is happening around the robot and are responsible for localization. Combining them with encoders in the wheels returns precise information about the change in position and the current position in space ❗ 💡These are two, independent "circuits" using the same terminal devices (scanners) 🎥 As you can see in the video, the viewing area of the scanners overlap which gives a full 360 degrees of "attention" - in addition, the scanners perfectly reproduce the shape of the elements around the robot ✅ 📜 Statement : Scanners are a key component in robotics. 🤖 PS: A video piece about scanners from SICK Sensor Intelligence ProCobot #robotics #automation #safety #mobileindustrialrobots #technologyfriday

72

2 Comments

Robotics has always been about thinking out of the box.

15

𝐖𝐡𝐚𝐭 𝐈𝐬 𝐓𝐡𝐞 𝐅𝐮𝐭𝐮𝐫𝐞 𝐎𝐟 𝐑𝐨𝐛𝐨𝐭𝐢𝐜 𝐕𝐢𝐬𝐢𝐨𝐧? The origins of computer vision and robotics trace back to the inception of artificial intelligence at the Dartmouth Conference in 1956. Initially termed machine vision, now known as robotic vision, it held promise for addressing key challenges in robotics, such as understanding the environment for planning and motion control. However, early endeavors encountered obstacles; cameras were bulky, computers lacked power, and obtaining quality 3D data was arduous. While stereo vision and LIDARs emerged in the 1980s, it wasn't until the early 1990s that the adaptation of industrial sensors as LIDARs revolutionized robotic vision. The majority of researchers working on three focus areas: 𝐑𝐨𝐛𝐮𝐬𝐭𝐧𝐞𝐬𝐬 The efficacy of a computer vision system hinges on its performance across diverse real-world conditions. Factors like lighting variations, shadows, and occlusions profoundly impact system reliability, posing a major obstacle for robotics. For robots, reliant on computer vision, true utility arises when systems deliver precise data across varied scenarios and with real-world objects. This is pivotal for tasks such as object recognition, tracking, navigation, and scene comprehension. 𝐑𝐞𝐚𝐥-𝐭𝐢𝐦𝐞 𝐏𝐞𝐫𝐟𝐨𝐫𝐦𝐚𝐧𝐜𝐞 Real-time processing entails the ability of a computer vision system to analyze incoming data on the fly and provide instantaneous results. This poses a significant challenge for robotics, especially for autonomous systems requiring up-to-the-moment information for decision-making and task execution. In robotics, swift real-time performance is indispensable for tasks like object recognition, tracking, navigation, and scene comprehension, enabling robots to operate effectively in dynamic environments. 𝐂𝐨𝐦𝐩𝐮𝐭𝐚𝐭𝐢𝐨𝐧𝐚𝐥 𝐄𝐟𝐟𝐢𝐜𝐢𝐞𝐧𝐜𝐲 Computational efficiency measures how adeptly a computer vision system processes data with speed and accuracy. For robotics, this poses a critical challenge as efficient information processing is fundamental for robots to make informed decisions and execute tasks seamlessly. Efficient computation is paramount in robotics for tasks such as object recognition, tracking, navigation, and scene comprehension, facilitating agile and effective robot operations. ------- Software and Hardware continues to get better and better, so researchers are making great strides in these areas. I am looking forward to hearing what the latest robotic vision breakthroughs are in the pipeline are at #ICRA2024. What do you think the future of robotic vision will be?

62

2 Comments

#AI #ComputerVision and #Perception enhance our robotics systems, making them more realistic and effective. Just an hour before bed, I was having fun with a pre-trained model with my test script towards the world’s largest crossing, Shibuya crossing 🌆 🇯🇵. It could more interesting for analyzing human behavior and traffic flow in dynamic urban environments. It’s crazy and fun 😃 #robotics #AI #research #startup #business #japan #tokyo #linkedin #techvoice #topvoice #japan #bangladesh #USA

5

Mastering Robot Code Structures: Learn the Basics of Labeling and Jumping in Programming In this video, we delve into the fundamentals of labeling and jumping in programming, using examples and explanations to help you understand how to structure your code effectively. Follow along as we explore pick positions and labels to optimize your coding workflow. #ProgrammingBasics #CodeStructures #LabelingAndJumping #OptimizedCoding #CodingWorkflow #CodeOptimization #CodingFundamentals #CodeStructureTips #ProgrammingTutorial #CodingTips

5

C𝗼𝗺𝗺𝗶𝘀𝘀𝗶𝗼𝗻 𝗯𝗲𝗳𝗼𝗿𝗲 𝗱𝗲𝗽𝗹𝗼𝘆𝗶𝗻𝗴 𝘁𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝘆. Using virtual reality with NVIDIA Omniverse. Ever since my first use of 3D simulation software back in 2009. I have dreamt of the day that we could commission automation equipment in a virtual world. That software was way ahead of its time. It had 3D graphics and full physics calculations. I could test how a box would travel up an incline. Unfortunately, the computer power wasn’t good enough to do an entire operation. ...15 years later both software and hardware have made the dream a reality. We can fully test equipment and even entire systems in a virtual environment. For a fraction of the cost to deploy real equipment, your system can be tested and verified. These are some of the benefits: ✅Inefficient designs can be identified and corrected at a minimal cost. ✅The rate of the system can be simulated before final purchase. ✅Change management and operator training can be done in a virtual world before equipment is installed. ✅Your existing operation will have minimal disruption. ✅On-site commissioning time and expenses can be reduced significantly. If you have ever been part of an automation project you understand that these benefits are pretty noteworthy. ===== Let’s take the risk out of automation. 📩DM me or schedule a meeting in the "featured" section of my profile to test your automation before it's installed. I deploy robotics and other technology to make warehouses safer and more efficient. I am your "easy button". ♻️Repost if you found this useful.♻️ #Oneiro #Automation #DigitalTwin #Omniverse *Video credit to FS Studio and Third Wave Automation

131

31 Comments

🆓 Degrees Of Freedom 🆓 📃 From definition: Degrees of freedom (DoF) refer to the ways in which an object can move or change in a spatial arrangement. 🤖 This is an integral part of robotics. Most robotic arms have 6 DoF, so the TCP point can change its position for all the possibilities shown in the video. 💡 However, there are robotic solutions that have fewer degrees of freedom - such as the SCARA-type robot, which does not have the ability to rotate the TCP point. 🔥 The degrees of freedom are directly linked to the "peculiar" positions of the arms - at which the arm is unable to make movements (usually linear). Peculiar positions will be characterized by such a kinematic arrangement of the arm/joints in which one (or more) of the DoF is lost. Enjoy the video from #Alex4D 🎥 #robotic #technology #kinematics

64

5 Comments

Tactile feedback is essential for understanding the dynamics of both rigid and deformable objects in many manipulation tasks, such as non-prehensile manipulation and dense packing. Researchers have introduced 𝐑𝐨𝐛𝐨𝐏𝐚𝐜𝐤, a framework combining visual and tactile sensing through a neural, tactile-informed dynamics model to enhance robotic manipulation. RoboPack employs a recurrent graph neural network to estimate object states, including particles and object-level latent physics information, from historical visuo-tactile observations and make future state predictions. This model, learned from real-world data, enables robots to solve downstream tasks using model-predictive control. The approach was demonstrated on a real robot equipped with a compliant Soft-Bubble tactile sensor. The robot tackles non-prehensile manipulation and dense packing tasks and infers objects' physics properties from direct and indirect interactions. Remarkably, the model was trained on only 30 minutes of real-world interaction data per task, yet it can perform online adaptation and make touch-informed predictions. RoboPack is showing promise over previous learning-based and physics-based simulation systems through extensive evaluations in long-horizon dynamics prediction and real-world manipulation. Congrats to the team, and best of luck with the future work. 📝 Research Paper: https://lnkd.in/e2Ng3VDA 📊 Project Page: https://lnkd.in/e-V3buQv #robotics #research

31

1 Comment

🆕 A Humanoid Robot in Germany - Neura 4NE-1 ! 🤖 Neura Robotics has unveiled their humanoid robot, 4NE-1, showcasing its capabilities in tasks like ironing and moving boxes. This announcement coincides with Nvidia's launch of new tools for humanoid development at the Siggraph conference. Neura is among the select companies granted early access to Nvidia’s advanced robotics tools, positioning it alongside industry leaders like Boston Dynamics and Figure. Neura's CEO, David Reger, emphasized the potential of combining Neura's cognitive robotics with Nvidia’s computing power to accelerate the evolution of humanoid robotics. Credits: Neura Robotics TechCrunch Source: https://lnkd.in/gfwryNgE https://lnkd.in/gA8sZ38V

13

#TGIF, Everybody! Have you ever been to a robotics competition and wondered why the robots were made to do certain things? Well, it may have been to test a new robot standard or see how far the technology can go in a certain application. From the Amazon Pick Challenge to the National Institute of Standards and Technology (NIST) Quadruped Challenge, these challenges on the surface are fun through promoting teamwork and teaching students to think outside the box. However, they are also helping Standards Development Organizations (SDOs) to see if a new standard is ready for the real world or if an existing standard needs updating, based on the outcomes of the challenge. Read more about all of this in my latest Six Degrees of Robotics column here: https://lnkd.in/eNQ5y55V

40

GTSAM is one of the most efficient and versatile frameworks for many optimization and estimation problems. By using factor graphs to represent optimization problems, GTSAM excels in handling the nonlinearities often encountered in robotics and computer vision. Its state-of-the-art nonlinear optimization techniques are crucial for achieving accurate and reliable results in these applications. For anyone working with localization and aiming to understand and build mapping or localization capabilities, GTSAM is an invaluable tool. It offers APIs in both C++ and Python, making it suitable for applications such as Structure from Motion (SFM), sensor fusion, navigation, trajectory optimization, and pose estimation. To know more about GT SAM, we made a basic and easy to understand tutorial on factor graph and GT SAM. Link - https://lnkd.in/dRgsMUBG #gtsam #localization #SLAM #Mapping #Robotics #Research

2

🚨IROS 2024 Paper Alert 🚨 ➡️Paper Title: Learning Variable Compliance Control From a Few Demonstrations for Bimanual Robot with Haptic Feedback Teleoperation System 🌟Few pointers from the paper 🎯Automating dexterous, contact-rich manipulation tasks using rigid robots is a significant challenge in robotics. Rigid robots, defined by their actuation through position commands, face issues of excessive contact forces due to their inability to adapt to contact with the environment, potentially causing damage. 🎯While compliance control schemes have been introduced to mitigate these issues by controlling forces via external sensors, they are hampered by the need for fine-tuning task-specific controller parameters. Learning from Demonstrations (LfD) offers an intuitive alternative, allowing robots to learn manipulations through observed actions. 🎯In this work, authors have introduced a novel system to enhance the teaching of dexterous, contact-rich manipulations to rigid robots. Their system is twofold: firstly, it incorporates a teleoperation interface utilizing Virtual Reality (VR) controllers, designed to provide an intuitive and cost-effective method for task demonstration with haptic feedback. 🎯Secondly, they presented Comp-ACT (Compliance Control via Action Chunking with Transformers), a method that leverages the demonstrations to learn variable compliance control from a few demonstrations. 🎯Their methods have been validated across various complex contact-rich manipulation tasks using single-arm and bimanual robot setups in simulated and real-world environments, demonstrating the effectiveness of their system in teaching robots dexterous manipulations with enhanced adaptability and safety. 🏢Organization: The University of Tokyo, OMRON SINIC X 🧙Paper Authors: Tatsuya Kamijo, Cristian Camilo Beltran-Hernandez, Masashi Hamaya 1️⃣Read the Full Paper here: https://lnkd.in/gu6Y8HaB 🎥 Be sure to watch the attached  Demo Video -Sound on 🔊🔊 Find this Valuable 💎 ? ♻️REPOST and teach your network something new Follow me 👣, Naveen Manwani, for the latest updates on Tech and AI-related news, insightful research papers, and exciting announcements.

20

2 Comments

🤖Graph-based Multi-sensor Fusion for Consistent Localization of Autonomous Construction Robots (Sup.)🤖 - Method for multimodal sensor fusion for robot state-estimation. - Graph-based prediction-update loop for consistent state estimates. - Enables operation of construction robots in diverse environments. - Dual-graph design for switching between optimization problems. - Validated on Menzi Muck walking excavators in real-world tests. Video - https://lnkd.in/e-eqQRaf Paper - https://lnkd.in/er7vmhGd Code - https://lnkd.in/e4f-2ZpZ #robotics #localisation

84

9 Comments

🤖 Robotics vs Biology 🐔 The stability of the chicken's head, its balance control system is very impressive! ⬇ 💡 In robotics we call it fixed TCP - the whole arm moves by changing its kinematic position keeping the TCP point in the same position (X,Y,Z). 🔥 In Universal Robots we can use both manually fixed tcp in freedrive or when moving in simple mode using arrows. 💁‍♂️ This helps a lot in situations where we need a stable TCP point, and different arm positions such as: ✅ Quality control, ✅ Assembly of small parts or ✅ Welding  Enjoy! #robotics #programming #tcp

358

15 Comments

Collaborative robot from KUKA as the steering wheel? Calm down, it's a very old concept from 2015 presented by HARMAN International based on the electrically-driven BMW i3, back then expected to be the next generation of autonomous urban driving. Economically absurdly unreasonable, yet satisfying to watch! 🦾 Imagine the 'let me drive!' arguments... 𝗔𝗟𝗟 𝗦𝗼𝗰𝗶𝗮𝗹 𝗠𝗲𝗱𝗶𝗮 𝗣𝗮𝗴𝗲𝘀 & 𝗖𝗼𝘂𝗿𝘀𝗲𝘀 https://lnkd.in/dH2k8u3i 𝐅𝐨𝐫 𝐀𝐥𝐥 𝐈𝐧𝐝𝐮𝐬𝐭𝐫𝐢𝐚𝐥 𝐑𝐨𝐛𝐨𝐭 𝐚𝐧𝐝 𝐏𝐋𝐂 𝐜𝐨𝐮𝐫𝐬𝐞𝐬 ( 𝐑𝐨𝐛𝐨𝐭𝐢𝐜𝐬 - 𝐀𝐮𝐭𝐨𝐦𝐚𝐭𝐢𝐨𝐧 𝐚𝐧𝐝 𝐀𝐈 𝐄𝐝𝐮𝐜𝐚𝐭𝐢𝐨𝐧 & 𝐓𝐫𝐚𝐢𝐧𝐢𝐧𝐠 ); https://lnkd.in/dGXHRA9G To access more 𝑨𝒑𝒑𝒍𝒊𝒄𝒂𝒕𝒊𝒐𝒏𝒔 and 𝑻𝒓𝒂𝒊𝒏𝒊𝒏𝒈𝒔 about 𝗥𝗼𝗯𝗼𝘁𝗶𝗰𝘀 𝗮𝗻𝗱 𝗔𝘂𝘁𝗼𝗺𝗮𝘁𝗶𝗼𝗻, you can "𝗙𝗼𝗹𝗹𝗼𝘄" from the link below. 🔔 𝗙𝗼𝗹𝗹𝗼𝘄 𝗣𝗲𝗿𝘀𝗼𝗻 : https://lnkd.in/dwhXFSxD ( Onur Sezgin 🤖 ) 🔔 𝗙𝗼𝗹𝗹𝗼𝘄 𝗣𝗮𝗴𝗲 : Robotics - Automation - Machine Learning and AI Trainings & Applications 🔔 𝗙𝗼𝗹𝗹𝗼𝘄 𝗚𝗿𝗼𝘂𝗽 : https://lnkd.in/d2_jFkjD 🔔 𝐖𝐡𝐚𝐭𝐬𝐚𝐩𝐩 𝐂𝐨𝐦𝐦𝐮𝐧𝐢𝐭𝐲 / 𝐆𝐫𝐨𝐮𝐩 𝐋𝐢𝐧𝐤 : https://lnkd.in/dqB-GHFa 🔔 𝐓𝐞𝐥𝐞𝐠𝐫𝐚𝐦 𝐂𝐨𝐦𝐦𝐮𝐧𝐢𝐭𝐲 / 𝐆𝐫𝐨𝐮𝐩 𝐋𝐢𝐧𝐤 : https://lnkd.in/dFqRW56q 🔔 𝗜𝗻𝘀𝘁𝗮𝗴𝗿𝗮𝗺 𝗣𝗮𝗴𝗲 : https://lnkd.in/d-rcY8Cw 🔔 𝗙𝗮𝗰𝗲𝗯𝗼𝗼𝗸 𝗣𝗮𝗴𝗲 : https://lnkd.in/dDeMmAQn 🔔 𝗧𝘄𝗶𝘁𝘁𝗲𝗿 (𝗫) 𝗣𝗮𝗴𝗲 : https://lnkd.in/dCbwZMph 🔔 𝗬𝗼𝘂𝘁𝘂𝗯𝗲 𝗣𝗮𝗴𝗲 : https://lnkd.in/d4NNhdkw #robotics #robots #robot #robotica #cobot #automation #ai #manufacturing #innovation #technology #Cobots #engineering #IndustrialAutomation #Technology #Mechatronics #openai #iot #innovations #technologies #electricity #ai #artificialintelligence #tech #future #programming #coding #software #machinelearning #engineer #stem #datascience #electric #computerscience #machine #industry #futuretech #softwaredeveloper #mechanical #newtechnology #cyber #nasa #space #arduino #raspberrypi #arduinoproject #diy #education #training #DeepLearning #AI #Science #Drones #Cloud #Startups #DeepMind #Industrial #CNC #3DPrinting #Manufacturing #MachineVision #HumanoidRobots #datascience #pythonprogramming #artificialintelligenceai #artificialintelligencefacts #artificialintelligenceart #pythons #artificialinspiration #deeplearning #codingisfun #technologynews #aiforgood #shorts

250

3 Comments

Mobile Robotics Today!!! For a mobile robotics designer, he must consider these two issues simultaneously when designing the locomoting mechanism of a wheeled robot. Why do wheel type and wheel geometry matter? three fundamental characteristics of a robot are governed by these choices: maneuverability, controllability and stability. 1. Maneuverability: This refers to the ability of a robot to change its position and orientation efficiently and effectively . It has several key aspect to be considered; turning radius, agility, navigation and precision. In general, a robot with high maneuverability is capable of navigating in a tight environment. 2. Controllability: This is the extent to which a robot's movements can be directed and modified by an operator or control system. Key aspects include; responsiveness, predictability, ease of control and feedback systems. In general, a robot with high controllability allows for precise and reliable operation, making it suitable for tasks that require fine control, such as medical robotics or delicate manufacturing processes. 3. Stability: This is the ability of a robot to maintain itself both statically and dynamically in it intended path of orientation without tipping over or deviating uncontrollably. key aspects to take note; static and dynamic stability, resilience to disturbance and balance mechanisms. ~ LEVON

7

1 Comment

#TGIF Everybody! Last week, my column in Six Degrees of Robotics about humanoids hit a nerve with some. While I agree that humanoid robots will address some critical use cases, the odds of them becoming the dominant form factor for #robotics going forward are very high against them in my opinion. One of the reasons for this is that currently, most people don't prefer robots that look human over those that don't. One only has to ask, if given a choice, would they rather have C-3PO or R2-D2? The vast majority pick the non-human-looking of the two over and over again. So, for this week's column, I look more at that preference and how it could be exactly the data point we need to help humanoids succeed by putting them in industries where people may prefer C-3PO over R2-D2. Read it here: https://lnkd.in/dXW5FeB2

29

4 Comments

Mobile robots and manipulation! 🕹️ Cooperative mobile manipulation is a crucial area in robotics, focusing on enabling robots to work together on tasks that require collaboration, much like humans do. This field addresses challenges such as transporting heavy or unwieldy objects in complex environments. To address these challenges, researchers from the Computational Robotics Lab - ETH Zürich have developed a multi-robot, bi-level optimization system based on direct transcription for wheeled mobile manipulation. This innovative approach utilizes static forces calculated for stability objectives at a lower level to inform wheeled trajectory planning at a higher level, resulting in effective planning and safer execution. Additionally, this system has been integrated with a previously developed Mixed-Reality interface! Congrats to authors Florian Kennel-Maushart, Stelian Coros! Tech stack and partners: Universal Robots, Clearpath Robotics by Rockwell Automation, Microsoft & Hilti Group! 🦾 ~~~ ♻️ Repost to help 1 robot find a new workplace. 📬 Subscribe to my newsletter to never confuse a robot and a cobot: https://lnkd.in/dzVGBKSc

425

16 Comments