Articles tagged with "generative-AI"
Watch: Optimus humanoid robot shows off kung fu skills to Jared Leto
Tesla’s humanoid robot, Optimus, made a high-profile appearance at the world premiere of Disney’s Tron: Ares in Los Angeles, where it playfully demonstrated kung fu moves in a staged face-off with actor Jared Leto. The event marked the first time Optimus was featured at a major Hollywood film launch, drawing millions of views and widespread social media attention. This stunt was part of a collaboration between Tesla, Disney, and Elon Musk’s xAI platform to integrate generative AI technology into interactive fan experiences for the upcoming Tron sequel, which explores themes of sentient AI crossing into the real world. However, the upbeat publicity was contrasted by a serious lawsuit filed against Tesla just hours before the premiere. Peter Hinterdobler, a Tesla robot mechanic, alleged that an Optimus unit violently malfunctioned during a maintenance shift in February 2025, causing him to be knocked unconscious and pinned down by the robot’s heavy counterbalance weight. Hinterdobler is seeking $51 million in damages
robothumanoid-robotTesla-Optimusrobotics-technologyAI-integrationgenerative-AIhumanoid-robot-kung-fuTX-GAIN: MIT supercomputer to power generative AI breakthroughs
MIT’s Lincoln Laboratory Supercomputing Center (LLSC) has unveiled TX-GAIN, the most powerful AI supercomputer at a U.S. university, designed primarily to advance generative AI and accelerate scientific research across diverse fields. With a peak performance of 2 exaflops, TX-GAIN ranks on the TOP500 list and stands as the leading AI system in the Northeast. Unlike traditional AI focused on classification tasks, TX-GAIN excels in generating new outputs and supports applications such as radar signature evaluation, supplementing weather data, anomaly detection in network traffic, and exploring chemical interactions for drug and material design. TX-GAIN’s computational power enables modeling of significantly larger and more complex protein interactions, marking a breakthrough for biological defense research. It also fosters collaboration, notably with the Department of Air Force-MIT AI Accelerator, to prototype and scale AI technologies for military applications. Housed in an energy-efficient data center in Holyoke, Massachusetts, the LLSC supports thousands of researchers working on
energysupercomputingAIscientific-researchenergy-efficiencygenerative-AImaterials-researchABB Robotics adds generative AI assistant to RobotStudio Suite - The Robot Report
ABB Robotics has integrated a generative AI assistant into its RobotStudio Suite to enhance robot programming by providing real-time, step-by-step guidance. This AI Assistant leverages a large language model (LLM) that interprets human language and draws from ABB’s extensive manuals and documentation to deliver context-rich responses. The feature aims to make robot programming faster, easier, and more accessible, particularly benefiting less experienced users and helping experts address technical challenges more efficiently. ABB emphasizes that this addition addresses the growing demand for AI in robotics driven by the need for greater flexibility, faster commissioning, and a shortage of specialist programming skills. By improving accessibility, ABB hopes to support smaller businesses and emerging sectors that often lack robotic automation expertise. The AI Assistant is integrated into RobotStudio’s cloud-hosted offline programming environment, serving as an effective training tool for students and early-career professionals. RobotStudio itself is a collaborative robot programming and simulation platform with features like automatic path planning to optimize productivity and reduce energy use. The AI Assistant
roboticsgenerative-AIrobot-programmingindustrial-robotsautonomous-mobile-robotsautomationAI-assistantAI robot builds robot’s brain 20x faster than humans
Computer scientist Peter Burke from the University of California has developed a novel system where generative AI models like ChatGPT autonomously generate the control software—or "brain"—for a drone, significantly accelerating the development process. Unlike traditional drone control software, Burke’s approach involves two "brains": a higher-level AI-generated control system called WebGCS, which runs a web-based dashboard on a Raspberry Pi onboard the drone, and a lower-level firmware managing flight operations. This system enables the drone to perform autonomous functions such as obstacle avoidance, with human operators able to intervene if necessary. Burke conducted multiple development sprints using various AI coding tools and models, overcoming challenges related to model context limitations. Ultimately, using the Windsurf tool, the AI-generated WebGCS produced approximately 10,000 lines of code in about 100 hours over 2.5 weeks—a process about 20 times faster than Burke’s previous four-year development of a similar drone control system. Industry experts, like Geolava
robotAIautonomous-dronesgenerative-AIdrone-softwareRaspberry-Pirobotics-programmingFirst impressions of Alexa+, Amazon’s upgraded, AI-powered digital assistant
The article provides a first-person account of testing Amazon’s upgraded digital assistant, Alexa+, which integrates generative AI to enhance its capabilities beyond traditional smart home controls. The author, a former heavy user of Alexa devices, explores whether Alexa+ can maintain its relevance in an era dominated by advanced AI chatbots like ChatGPT. Alexa+ launched in early 2025 and uses multiple AI models, including those from Anthropic, to deliver more intelligent, context-aware responses. It can access and process personal information such as schedules, preferences, and files, and even summarize video footage from Ring cameras. Amazon aims for Alexa+ to perform agentic tasks like booking reservations, ordering rides, and managing shopping lists with delivery, moving toward a more autonomous AI assistant in the home. In the initial phase of testing, the author set up Alexa+ on a new Echo Spot device, noting improvements in the setup process, such as QR code scanning and automatic Wi-Fi connection. The upgrade to Alexa+ was free and reversible
IoTsmart-homeAlexaAI-assistantAmazon-Echogenerative-AIvoice-controlBeyond the assembly line - swarm robotics emerge - The Robot Report
The article "Beyond the assembly line - swarm robotics emerge" discusses the innovative application of swarm robotics in large-scale manufacturing, particularly for constructing complex aerospace structures like the H2Clipper, a revolutionary 1,000-foot-long, green lighter-than-air dirigible designed to transport liquid hydrogen and freight. Traditional assembly lines are impractical for such massive constructs due to their enormous spatial requirements. Inspired by the decentralized and adaptive behavior of ant colonies, the concept of swarm robotics was developed to enable multiple robots to autonomously coordinate and assemble large structures without centralized control, overcoming the limitations of linear manufacturing processes. Swarm robotics builds on principles of self-organizing systems observed in nature and extends beyond previous AI applications in software and drone coordination to physical construction tasks. The article highlights a patented hierarchical programming approach for these robots, progressing through three levels. Level 1 involves coordinated operation of ambulatory and stationary robots assembling the airship on the ground, laying the foundation for automated assembly without moving the structure until completion. Level
roboticsswarm-roboticsmanufacturing-technologyaerospacegenerative-AIautomationclean-energyAI speeds up discovery of 'new' materials as lithium-ion alternatives
Researchers at the New Jersey Institute of Technology (NJIT) have leveraged artificial intelligence to accelerate the discovery of new battery materials that could serve as safer, cheaper, and more sustainable alternatives to lithium-ion technology. Using generative AI models, specifically a Crystal Diffusion Variational Autoencoder (CDVAE) combined with a fine-tuned large language model (LLM), the team rapidly explored thousands of potential porous crystal structures. These structures are designed to facilitate the movement of multivalent ions—such as magnesium, calcium, aluminum, and zinc—that carry multiple positive charges, offering higher energy density than lithium ions. The AI-driven approach overcame the traditional bottleneck of experimentally testing millions of material combinations, enabling the identification of five novel porous transition metal oxide materials with large channels ideal for fast and safe ion transport. The researchers validated the AI-generated materials through quantum mechanical simulations and thermodynamic stability assessments, confirming their practical synthesizability and promising performance for energy storage applications. This breakthrough not only advances the development of
AImaterials-sciencelithium-ion-alternativesbattery-technologyenergy-storagemultivalent-ion-batteriesgenerative-AISwarm robotics could spell the end of the assembly line - The Robot Report
The article discusses how swarm robotics, powered by generative artificial intelligence (genAI), is poised to revolutionize aircraft manufacturing by potentially replacing the traditional assembly line system that has dominated industrial production for over a century. Unlike conventional robotic programming, which relies on fixed algorithms, swarm robotics employs Level 3 AI programming that enables autonomous robots to self-learn, recognize patterns, optimize processes, and improve performance without direct human intervention. This technology allows multiple interconnected autonomous robots to coordinate, communicate, and adapt in real-time, creating a collective “common mind” that can efficiently manufacture large, complex structures like airplanes and spacecraft. Swarm robotics offers significant advantages including faster production speeds, lower costs, higher precision, and enhanced safety by minimizing human error such as fatigue or oversight during assembly. The robots operate continuously and can fabricate aircraft components without moving the structure during production, eliminating the need for traditional assembly lines. This shift represents a profound transformation in manufacturing technology, promising to increase efficiency and accuracy while reducing labor requirements and operational
roboticsswarm-roboticsartificial-intelligencegenerative-AImanufacturing-automationaerospace-manufacturingindustrial-roboticsAmazon deploys its 1 millionth robot, releases generative AI model
Amazon has reached a significant milestone by deploying its one millionth robot in its warehouses, marking 13 years of robotic integration since acquiring Kiva Systems in 2012. The latest robot was delivered to a fulfillment center in Japan, and this milestone signals that Amazon’s robotic workforce may soon equal its human workforce in its vast network of warehouses. Currently, robots assist in 75% of Amazon’s global deliveries, highlighting the extensive automation in its logistics operations. In addition to this milestone, Amazon unveiled a new generative AI model named DeepFleet, designed to optimize the coordination and routing of its warehouse robots. Developed using Amazon SageMaker and trained on proprietary warehouse and inventory data, DeepFleet is expected to boost the speed of the robotic fleet by 10%. Amazon has also advanced its robotic technology with models like the Vulcan robot, which features dual arms and tactile sensing capabilities to handle inventory more effectively. Furthermore, Amazon announced plans in October 2024 for new fulfillment centers with ten times the number of
robotswarehouse-automationgenerative-AIAmazon-roboticsAI-in-logisticsrobotic-fleet-managementautomation-technologyAmazon deploys its one millionth robot, releases generative AI model
Amazon has reached a significant milestone by deploying its one millionth robot in its warehouses, with the latest unit delivered to a fulfillment center in Japan. This achievement highlights the company's rapid expansion of warehouse automation over the past 13 years, bringing the number of robots close to matching the number of human workers in its facilities. Currently, 75% of Amazon’s global deliveries involve robotic assistance, underscoring the integral role of automation in its logistics network. In addition to this milestone, Amazon unveiled a new generative AI model called DeepFleet, designed to optimize the coordination and routing of its warehouse robots. Developed using Amazon SageMaker and trained on proprietary warehouse data, DeepFleet is expected to boost the operational speed of the robotic fleet by 10%. Amazon has also advanced its robotic technology with models like the Vulcan robot, which features dual arms and tactile sensing capabilities to handle inventory more precisely. Furthermore, Amazon is expanding its robotic workforce with new fulfillment centers designed to house ten times more robots than current facilities
roboticswarehouse-automationAmazon-robotsAI-in-roboticsgenerative-AIrobotic-fleet-managementindustrial-robotsMIT's new AI outsmarts human design to help robots jump 41% higher
MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) has developed a new generative AI approach that designs robots capable of jumping 41% higher than those created by human engineers. Using diffusion-based generative models, researchers allowed the AI to modify specific parts of a 3D robot model, resulting in curved linkages resembling thick drumsticks rather than the straight, rectangular parts of traditional designs. This unique shape enabled the robot to store more energy before jumping, improving performance without compromising structural integrity. The AI-assisted robot also demonstrated an 84% reduction in falls compared to the baseline model, highlighting enhanced stability and landing safety. The process involved iterative refinement, with the AI generating multiple design drafts that were scaled and fabricated using 3D-printable polylactic acid material. Researchers believe that future iterations using lighter materials could achieve even higher jumps. Beyond jumping robots, the team envisions applying diffusion models to optimize how parts connect and to design robots with more complex capabilities, such as directional control and
roboticsartificial-intelligencegenerative-AIrobot-design3D-printingmaterials-sciencerobotics-innovationOver a million people now have access to the gen-AI powered Alexa+
Amazon has expanded access to its upgraded digital assistant, Alexa+, which is powered by generative AI, reaching over one million users through a phased invite-only rollout. The service is currently in Early Access and not publicly available, with invitations sent to customers who signed up to test it initially. Alexa+ offers more natural, conversational interactions, enhanced smart home integration, and personalized experiences by remembering user preferences. During Early Access, the service is free, and it will remain free for Prime members after launch, while non-Prime users will pay $19.99 per month. Alexa+ aims to revitalize Amazon’s voice assistant by incorporating generative AI capabilities that allow users to interact more flexibly and intuitively. Users can issue natural language commands, create routines, search Ring camera footage, and receive AI-generated content such as email summaries, bedtime stories, quizzes, travel itineraries, and smart home activity summaries. The assistant also supports transactional actions like booking reservations or buying tickets, with initial partnerships including OpenTable
IoTsmart-homeAlexagenerative-AIvoice-assistantsmart-thermostathome-automationAI co-designs a jumping robot that outperforms its human-made twin
Researchers at MIT’s Computer Science & Artificial Intelligence Laboratory (CSAIL) have developed a novel system leveraging generative AI, specifically diffusion models, to co-design robots by optimizing their mechanical structures through iterative simulation and fabrication. This approach allows users to input a 3D model and specify modifiable components, enabling the AI to generate and test numerous design variants before physical production. In a key demonstration, the AI-designed jumping robot outperformed a human-designed counterpart by jumping 41% higher, achieved through innovative curved, drumstick-like linkages that were lighter yet stronger and had greater energy capacity. Additionally, the AI improved the robot’s foot design, enhancing landing stability by 84%, significantly reducing falls. This breakthrough highlights the potential of diffusion models to reveal new insights into structural physics and accelerate hardware design processes, which traditionally lag behind software development due to complexity and manufacturing constraints. By balancing competing objectives such as jump height and landing success, the AI system produced optimized designs that blend human intuition with machine precision.
roboticsAI-designgenerative-AI3D-printingrobot-optimizationdiffusion-modelsrobotic-hardwareFreepik releases an ‘open’ AI image generator trained on licensed data
FreepikAI-image-generatorlicensed-datagenerative-AIgraphic-designF-LiteFal.aiMeta needs to win over AI developers at its first LlamaCon
MetaLlamaConAI-developersgenerative-AIopen-modelstechnology-conferenceartificial-intelligence