Industry Briefing
A single destination for timely, editor-curated robotics news from around the world.
Chinese AI company offers a new solution for physical AI in the uncertain trillion-dollar market.
In 2026, the field of physical AI is set to emerge as a transformative force, following a consensus reached by industry leaders at the CES in Las Vegas, where NVIDIA's CEO Jensen Huang heralded the arrival of "physical AI's ChatGPT moment." Over the past two years, significant advancements have been made in five key areas: brain models, imagination engines, training environments, ontology, and commercial ecosystems, laying the groundwork for real-world applications. In the first half of 2026, global investment in physical AI surged, with over $6.4 billion raised in just the first quarter, including notable funding rounds from AMI Labs and World Labs. The industry is witnessing a clear technological divergence, with three primary paths emerging: Visual Language Models (VLM), Visual Language Action (VLA), and world models. The anticipated future architecture for physical AI is expected to integrate VLA's decision-making capabilities with world models' predictive simulations. Despite the rapid growth, the competitive landscape remains uncertain, with various companies pursuing different strategies, including those focusing solely on VLA or world models, and others exploring hybrid approaches. The ultimate goal is to develop AI that can effectively navigate and understand the complexities of the physical world, moving beyond mere reactive capabilities to proactive, autonomous decision-making. As the physical AI market is projected to expand significantly, reaching an estimated $3.26 trillion by 2040, the industry faces the challenge of ensuring that technology translates into tangible business value. Companies like Om AI are pioneering innovative models that prioritize continuous perception and spatial understanding, aiming to redefine how AI interacts with its environment. The ongoing evolution of physical AI emphasizes the importance of real-world applications and the need for AI systems that can adapt and respond to dynamic physical spaces.
3 Questions: Beyond data-driven aesthetics
Alexandros Haridis, a recent graduate with a master's degree in 2017 and a PhD in 2022, is showcasing his work in a new exhibition at the Keller Gallery. The exhibition delves into the evolution of aesthetic judgment over the centuries and examines the role of design in rendering complex computational systems more accessible and understandable. By highlighting the intersection of art and technology, Haridis aims to foster a deeper appreciation for how design influences our perception of intricate data. The exhibition is set to attract visitors interested in both the historical context of aesthetic principles and contemporary design practices.
School of Architecture and Planning Architecture Computer science and technology Artificial intelligence Machine learning Algorithms
Zhipingfang Hits $2.8B Valuation as Brain-Like AI Era Dawns, Greater Bay Area's First Embodied AI Unicorn
Zhipingfang, a company specializing in embodied artificial intelligence, has successfully raised around 5 billion yuan in new funding. This significant financial boost has propelled the company's total valuation to over 20 billion yuan. The funding is expected to enhance the development of Zhipingfang's innovative NeuroVLA architecture, which is inspired by brain functions. The announcement of this funding round highlights the growing interest and investment in advanced AI technologies, reflecting the industry's potential for future growth and innovation.
Startups
Liquid AI's smallest model yet LFM2.5-230M beats models 4X its size at data extraction, can run 'anywhere'
Liquid AI, a company founded by former MIT computer scientists, has unveiled its latest AI language model, LFM2.5-230M, which is designed for efficient data extraction and local deployment on devices such as smartphones and laptops. Released today, this 230-million-parameter model is noted for its ability to run on various hardware platforms, outperforming larger models like Alibaba's Qwen3.5 and Google's Gemma 3 in specific benchmarks. Targeting developers and engineers, LFM2.5-230M operates under a dual-use commercial license, allowing free access for individuals and companies with annual revenues below $10 million, while larger enterprises must secure a paid agreement. The model distinguishes itself by utilizing the LFM2 architecture, enabling high inference speeds with a minimal memory footprint, making it suitable for edge computing. Liquid AI's launch reflects a broader industry shift towards architectural efficiency rather than sheer parameter counts, as major AI firms focus on models with hundreds of billions of parameters. The LFM2.5-230M is specifically tailored for lightweight data extraction tasks, allowing businesses to automate processes without relying on costly cloud services. In practical applications, the model has been successfully deployed in a humanoid robot, demonstrating its capability to process complex commands efficiently. Available immediately on platforms like Hugging Face, LFM2.5-230M aims to revolutionize how enterprises manage data extraction, moving away from traditional, rigid systems to more adaptable AI-driven solutions.
Technology
Why Does a Bank Need a Chief Scientist?
Prem Natarajan, formerly the head of Alexa AI at Amazon, has transitioned to the role of Chief Scientist at Capital One, a leading financial institution serving over 100 million customers. This move comes as the landscape of artificial intelligence (AI) research shifts from large tech companies to industry-specific applications, particularly in finance, where the challenges are more complex and require a nuanced understanding of customer needs and regulatory constraints. Capital One has long been recognized for its data-driven approach, having invested heavily in cloud technology to create a robust data ecosystem. This infrastructure supports innovative AI research aimed at solving real-world problems, such as real-time fraud detection and personalized customer interactions. Unlike many banks that view AI merely as a tool, Capital One is establishing a scientific community focused on developing impactful AI solutions. Natarajan emphasizes the importance of “destination-back thinking,” where the team envisions ideal customer experiences and works backward to identify necessary scientific advancements. This methodology, combined with a commitment to continuous learning and a unique cloud-first architecture, allows Capital One to tackle challenges that are often overlooked by traditional financial institutions. The bank's efforts have garnered recognition, with Capital One being ranked as a leader in AI talent and innovation, accounting for a significant portion of AI patents in the financial sector. Through partnerships with academic institutions and a focus on agentic AI systems, Capital One aims to enhance financial services for millions, positioning itself at the forefront of AI advancements in the industry.
Ai-research Agentic-ai Financial-services Tech-careers Type-sponsored Financial-technology
RoboScience launches Visics, a versatile embodied model for cross-ontology, cross-object, and cross-task applications.
On June 24, RoboScience, a company specializing in embodied intelligence, unveiled its self-developed Visics large model, introducing the innovative VLOA (Vision-Language-Object-Action) architecture. This announcement marks a significant advancement in the field, demonstrating the model's applications in real-world scenarios such as furniture assembly, dexterous grasping, and dynamic assembly lines. The current landscape of embodied intelligence lacks a universally accepted foundational representation unit, which hampers data collection, model learning, and the transfer of knowledge to new contexts. Traditionally, models have focused on replicating specific robotic movements tied to particular tasks, limiting their adaptability to new robots, objects, or environments. Founder and CEO Tian Ye highlighted three major challenges in robotic operations: poor generalization, difficulty in precise manipulation, and cumulative errors in long-range tasks. To address these issues, RoboScience has developed a new foundational representation unit from the ground up. The Visics model employs a dual-engine architecture, consisting of an embodied world model and a universal operation model, each operating independently. The embodied world model utilizes vast amounts of internet video data to learn the physical dynamics of objects, while the operation model translates object trajectories into actionable commands for robots. This layered design enhances the model's generalization capabilities across various robotic platforms and tasks. RoboScience's innovative approach also includes a high-precision simulation engine, RoboMirage, which, combined with automated video data annotation, significantly reduces data acquisition costs. The company aims to build a comprehensive dataset of over 1 terabyte of high-quality manipulation trajectories by 2026. Since its inception, RoboScience has garnered support from multiple investors and established research and production centers in major Chinese cities. The company plans to collaborate with various sectors, including retail and logistics, to standardize robotic products for industrial and commercial applications by the end of this year.
Qualcomm vs. Nvidia and drones vs. dogs
Qualcomm has made significant strides in the semiconductor industry, unveiling its ambitious data center chip roadmap during its annual investor day in New York on June 25, 2026. CEO Cristiano Amon highlighted the company's new AI accelerator platform and innovative chip architecture known as high-bandwidth compute (HBC), which aims to enhance AI processing by reducing data travel distances and energy consumption. This announcement comes amid a busy day for the tech sector, where Nvidia's CEO Jensen Huang reaffirmed the long-term demand for AI infrastructure, and Micron reported strong earnings, alleviating investor concerns about a potential "AI bubble." Qualcomm's focus on the China market is particularly noteworthy, as the country accounted for 46% of its revenue in 2025. Amon indicated that the company is designing chips tailored for Chinese customers while adhering to U.S. export controls. This strategic move aims to leverage Qualcomm's existing relationships with Chinese smartphone manufacturers to expand its data center business. Meanwhile, Nvidia's AI chips have seen a dramatic price increase in China's black market, driven by strong demand and U.S. export restrictions. The price of Nvidia's flagship DGX B300 server has surged to over 8 million yuan ($1.1 million), reflecting the ongoing challenges in accessing these sought-after technologies. In a separate development, Australian farmers are increasingly turning to drones and AI technologies for livestock management, potentially replacing traditional herding dogs. This shift highlights the evolving landscape of agricultural practices as new generations of farmers adopt innovative solutions to enhance efficiency in managing livestock.
IEEE Rolls Out Large Language Models Virtual Training Course
Large language models (LLMs) have transitioned from research labs to everyday use in engineering, significantly altering how digital infrastructures are developed and maintained. As technical professionals increasingly rely on LLMs for complex tasks—such as identifying vulnerabilities in source code and converting fragmented discussions into detailed specifications—the demand for expertise in this technology is surging. According to MarketsandMarkets, the LLM technology market is projected to grow by approximately 33% annually through 2030. To effectively utilize LLMs, engineers must move beyond basic interactions and understand the underlying transformer architecture that enables these models to process vast datasets simultaneously. This knowledge is crucial to mitigate risks associated with inaccuracies, often referred to as "hallucinations," and to ensure reliable performance in coding and data handling. Key advancements include integrating LLMs with application programming interfaces (APIs) for direct database connections, addressing hallucination issues through retrieval-augmented generation (RAG), and prioritizing data security by establishing private model instances. Additionally, LLMs automate repetitive tasks, allowing engineers to focus on higher-level design and problem-solving. To bridge the growing knowledge gap, IEEE has launched an online program titled "Large Language Models Demystified," designed to equip technical professionals with a deeper understanding of LLMs. The curriculum covers the evolution of AI technology, transformer architectures, and practical model-building exercises. Participants will earn professional development credits and a digital badge upon completion, enhancing their credentials in this rapidly evolving field. Organizations interested in training their teams can consult with IEEE for tailored enrollment options.
Ai Type-ti Education Ieee-educational-activities Large-language-models Ieee-products-and-services
NIO enhances smart driving education; Ren Shaoqing states innovation will reshape competition.
On June 18, NIO announced the rollout of its latest world model software across multiple vehicle platforms, including eight NT2.0 models, four NT2.5 models, and six NT3.0 models. This update allows NIO to run the same complex autonomous driving code on different generations of chips, addressing a common industry challenge where software updates were often limited to the latest hardware, leaving older vehicle owners at a disadvantage. The initiative stems from a long-term effort by NIO's team, led by Ren Shaoqing, who began exploring solutions in 2020. NIO developed an AI infrastructure that bridges gaps between different chip architectures, enhancing vehicle processing speeds with an AI compiler and automating deployment processes with AI agents. This innovation has significantly reduced deployment times from days to just a couple of hours. NIO's approach includes running the latest models in a "shadow mode" on production vehicles to gather valuable data without interfering with user driving. This data is used to train smarter models, creating a feedback loop that enhances the software's performance. The company has reported a substantial increase in its autonomous driving capabilities, attributing this to a shift in understanding the development cycle of physical AI. As the industry evolves, NIO has restructured its autonomous driving team to focus on foundational research and innovation, positioning itself to leverage advancements in large model technology and closed-loop reinforcement learning. The company aims to enhance its competitive edge by continuously improving its algorithms and data systems, ultimately striving for a more robust autonomous driving experience.
Graitec outlines AI roadmap focused on accountability in engineering and construction
Graitec, a software provider, has announced a comprehensive three-stage artificial intelligence strategy tailored for the architecture, engineering, construction, and operations (AECO) sector. The company emphasizes that the primary challenge facing the industry is not the creation of content through AI, but rather the reliability of these AI-generated results in practical applications. To address this concern, Graitec plans to integrate AI directly into various processes, including engineering, fabrication, and construction. This initiative aims to enhance trust and efficiency within the AECO industry, ultimately improving project outcomes.
Construction News Software AECO AI software architecture and engineering
AGIBOT Showcases Embodied AI Robots at VivaTech 2026 in Paris
AGIBOT unveiled its cutting-edge embodied AI robots during VivaTech 2026, held in Paris. The company conducted live demonstrations showcasing the robots' advanced interaction, locomotion, and manipulation abilities. Central to their presentation was the 'Three Intelligences in One' architecture, which underscores the evolution of humanoid robotics from experimental concepts to practical applications. AGIBOT also engaged with industry leaders to discuss the future trajectory of AI and robotics, highlighting the growing relevance of these technologies in various sectors.
Embodied AI Robotics Technology Innovation Humanoid Robotics
European Defense Firms Partner on Sensor-to-Interceptor Counter-Drone Network
European counter-drone companies Alpine Eagle and Origin Robotics have formed a partnership to develop an integrated counter-drone system. This collaboration, formalized through a memorandum of understanding, aims to combine Origin Robotics’ BLAZE interceptor with Alpine Eagle’s Sentinel architecture. The initiative seeks to enhance air defense capabilities by creating a unified network that incorporates detection, command-and-control, and interception technologies. This strategic alliance reflects the growing need for advanced solutions to counter drone threats in Europe.
Anti-drone technology Applications C-UAS Defense defense Drone News
Zhi Ping Fang launches NeuroVLA, a brain-like embodied intelligence system.
Recently, Zhi Ping Fang unveiled NeuroVLA, an innovative brain-inspired embodied intelligence system, marking a significant advancement in robotics. This system features a unique three-tier architecture that mimics the human brain, comprising the cortex for semantic understanding and task planning, the cerebellum for high-frequency motion coordination and dynamic correction, and the spinal cord for rapid motion execution and safety reflexes. Experimental results indicate that NeuroVLA can reduce robotic motion jitter by over 75% and achieve reflex responses within 20 milliseconds following a collision, all while significantly lowering system power consumption. This breakthrough aims to enhance the efficiency and safety of robotic operations, positioning NeuroVLA at the forefront of intelligent robotics technology.
The Enlightenment World Model tops evaluations in RoboTwin 2.0 and other embodied intelligence tests.
Recently, DaXiao Robotics announced that its Kairos world model has achieved top rankings in several global evaluations focused on embodied intelligence, including RoboTwin 2.0, LIBERO-Plus, WorldModelBench Robot, and DreamGen. This model employs an integrated architecture that combines multimodal understanding, generation, and prediction. In a significant move for the industry, DaXiao Robotics has made the Kairos model open-source, allowing broader access and collaboration in the field of AI-driven video generation and state prediction.
indie Launches Edge AI SoC to Power Smarter Perception Systems for Automotive and Humanoids
indie, an automotive solutions innovator based in Aliso Viejo, CA, has announced the launch of its next-generation edge AI system-on-chip (SoC), the iND881, designed to enhance smart camera technology for automotive and robotic applications. Unveiled on June 10, 2026, the iND881 integrates an AI compute engine with indie’s advanced low-latency multi-camera image signal processor (ISP), providing an efficient solution for developers. The iND881 is engineered for low power consumption and real-time responsiveness, featuring a Neural Processing Unit (NPU), a versatile Digital Signal Processor (DSP), and a quad-core ARM Cortex-A53 CPU. This architecture is tailored for demanding edge perception tasks, making it particularly beneficial for advanced driver assistance systems, including driver and occupant monitoring and smart mirrors with blind-spot detection. In addition to automotive applications, the iND881 supports robotics and physical AI automation, facilitating accurate sensing and navigation for autonomous mobile robots. The device's capabilities include multi-channel video compression, high-dynamic-range ISP, and compatibility with various sensor modalities such as infrared and LiDAR, ensuring robust performance in complex environments. The iND881 is ASIL-B compliant and automotive qualified, currently available for sampling. It will be showcased at the upcoming AutoSens and InCabin USA 2026 events. Fred Jarrar, indie's senior vice president, emphasized that the launch not only expands their product portfolio but also positions indie as a comprehensive solutions provider in the edge AI market.
Beyond Dexterity: Why Contact May Define the Next Era of Robotics
At the 2026 IEEE International Conference on Robotics (ICRA) in Vienna, AGILINK showcased a captivating demonstration of robotic dexterity by creating a balloon dog, which drew significant attention from attendees. This seemingly playful task is recognized in the robotics community as a complex manipulation challenge due to the balloon's lightweight and highly deformable nature. The demonstration highlighted the intricate balance between motion and contact intelligence, essential for successful robotic manipulation. AGILINK's approach involved mapping the actions of professional balloon artists to robotic hands, allowing the robot to learn both successful manipulation sequences and recovery strategies during failures. This dual focus on motion and contact intelligence is crucial, as maintaining stable interaction with the balloon is as important as executing the correct sequence of actions. In conjunction with the balloon dog demonstration, AGILINK introduced the OmniHand 3 Ultra-M, a dexterous robotic hand designed to enhance contact intelligence through advanced sensing and faster response capabilities. The hand features 20 active degrees of freedom and a direct-drive architecture, enabling precise force regulation and tactile sensing across its surface. The significance of these advancements extends beyond balloon animals, addressing broader challenges in robotics related to unstable and deformable interactions, such as delicate assembly and household tasks. As robotics research increasingly prioritizes interaction dynamics, AGILINK's innovations may pave the way for more effective manipulation in unpredictable real-world environments.
Humanoid-robots Physical-ai Dexterous-hands Direct-drive-actuation Robotic-manipulation Reinforcement-learning
Cognizant Launches Sovereign Physical AI Platform-As-A-Service
Cognizant has unveiled a new sovereign Physical AI Platform-as-a-Service aimed at assisting enterprises in the integration and management of autonomous systems, industrial equipment, and AI-driven operations through a unified software layer. This innovative platform, developed on the Cognizant Intelligence Spine architecture, is designed to seamlessly connect various technologies, including sensors, cameras, robots, and digital twins. By providing a comprehensive solution, Cognizant seeks to enhance operational efficiency and streamline the management of complex industrial environments. The launch reflects the company's commitment to advancing AI capabilities within the enterprise sector, enabling businesses to leverage cutting-edge technology for improved performance and productivity.
AI AI Funding & Investment AI Use Cases Robotics Cognizant Energy
UMass Amherst Researchers Developing AI Architecture That Uses a Fraction of the Energy Required by Today’s AI Systems
Researchers at the University of Massachusetts Amherst have unveiled a groundbreaking artificial intelligence architecture aimed at significantly lowering the energy consumption of advanced AI systems while maintaining their learning capabilities. This innovative approach, inspired by brain function, was developed with funding from the U.S. National Science Foundation and the Air Force Office of Scientific Research. By mimicking the efficiency of the human brain, the new architecture seeks to address the growing energy demands associated with AI technologies, which have raised concerns regarding sustainability and environmental impact. The research, which highlights the potential for more eco-friendly AI solutions, could pave the way for advancements in various fields reliant on artificial intelligence, ultimately promoting a more sustainable future for technology.
AI AI Research & Advances Robotics architecture energy consumption Research
KONGSBERG and DRASS Announce Strategic Partnership
KONGSBERG and DRASS have announced a strategic partnership aimed at enhancing the development of advanced underwater solutions. This collaboration will facilitate the exchange of technologies, systems, payloads, and operational architectures, leveraging the complementary industrial strengths and technical expertise of both companies. By combining their operational experience, KONGSBERG and DRASS seek to innovate and improve underwater capabilities, addressing the growing demands in this sector. The partnership marks a significant step forward in the pursuit of cutting-edge solutions for underwater operations.
kongsberg drass strategic partnership
Nvidia’s AI Hardware Comes to Windows in RTX Spark PCs
At Computex 2026 in Taipei, Taiwan, Nvidia unveiled its highly anticipated RTX Spark superchip for Windows PCs, marking a significant development in the tech industry. This announcement, which comes a year later than initially expected, was made in collaboration with Microsoft, which introduced two new devices powered by the RTX Spark: the Surface Laptop Ultra and the Surface RTX Spark Dev Box. Major PC manufacturers, including Asus, Dell, Lenovo, HP, and MSI, also showcased their Windows PCs featuring the new chip. The RTX Spark is based on Nvidia's Blackwell GB10 architecture, boasting 20 Arm CPU cores, 6,144 GPU cores, and support for up to 128 gigabytes of LPDDR5X memory. While the chip is designed to consume less power than its predecessor, the DGX Spark, it is expected to maintain strong performance, particularly in gaming and professional applications. Analysts suggest that Nvidia's established presence in the GPU market, with over 90% share, will enhance the software ecosystem for RTX Spark, setting it apart from previous attempts by Qualcomm and Microsoft with their AI-focused Copilot+ PCs. As Nvidia and Microsoft aim to position RTX Spark as a viable alternative to traditional x86 chips from Intel and AMD, they face the challenge of proving its effectiveness as a general-purpose PC. The launch is seen as a strategic move to leverage AI capabilities while appealing to both creators and gamers, with Nvidia emphasizing the importance of robust software support alongside hardware advancements. RTX Spark desktop workstations are expected to be available in the third quarter of 2026, further expanding the potential applications of this new technology.
Nvidia Pcs Windows Arm Ai-hardware
Software is ‘the biggest bottleneck to robotics innovation’, says BlackBerry QNX report
QNX, a division of BlackBerry, has unveiled its latest research study, the Inside the Robot: Architecture Benchmark Report, which explores the evolving landscape of robotics development. The report highlights the shift towards software-driven and AI-enabled systems that are increasingly integrated into workplaces and everyday life. Conducted through a survey of 1,000 developers globally, the research aims to shed light on the current trends and challenges faced in the robotics sector. This initiative reflects QNX's commitment to understanding and advancing the role of robotics in modern society, emphasizing the importance of collaboration between humans and machines. The findings are expected to inform future developments in the field and guide industry stakeholders in adapting to these transformative changes.
Features Robotics Software ai robotics automation news Autonomous robots
New Server Hopes to Break Through AI’s “Memory Wall”
Majestic Labs, an AI hardware startup, is addressing the memory limitations of large language models (LLMs) with its upcoming server, Prometheus, set to launch in 2027. This innovative server will feature up to 128 terabytes of memory, significantly surpassing the capabilities of Nvidia’s current offerings. Co-founder Sha Rabii emphasizes that this substantial memory increase will enhance performance and efficiency, particularly as models grow larger. Prometheus employs a unique DRAM-centric architecture, utilizing LPDDR6 memory and a proprietary memory interface with miniature copper cables that allow for greater memory placement flexibility. This design aims to overcome the “memory wall” that hampers LLM performance, providing a memory bandwidth of up to 25.6 terabytes per second. To complement its memory capabilities, Prometheus will incorporate the Ignite AI processing unit, which combines ARM application cores with RISC-V vector and tensor cores on a single chip. This integration allows for seamless handling of LLM inference tasks without the need for processor handoffs. Majestic Labs is also focused on ensuring compatibility with existing AI frameworks like PyTorch and OpenAI’s Triton, allowing customers to run their models without modifications. The server, designed in compliance with the Open Compute Project, will be modular, enabling future memory upgrades. Despite the advanced technology, Majestic Labs aims to offer competitive pricing by leveraging DRAM instead of more expensive high-bandwidth memory. Rabii claims that this approach could reduce customer capital expenditures and power consumption significantly, potentially by 10 to 50 times, depending on the workload.
Memory Server Ai-accelerators Performance
Study: Regulation and safety hinder robotics projects in Germany
QNX, a division of BlackBerry, has released findings from its latest study titled "Inside the Robot: Architecture Benchmark Report." The report highlights that regulatory challenges and safety concerns are hindering robotics projects in Germany. This study sheds light on the current state of the robotics industry, emphasizing the need for clearer regulations and enhanced safety measures to facilitate innovation and development in this sector. The findings are particularly relevant as Germany seeks to advance its technological capabilities in robotics amidst growing global competition.
Allgemein Newsarchiv
Nvidia jumps into PCs with new Arm-based chip debuting in laptops from Microsoft, Dell, HP
Nvidia CEO Jensen Huang announced the launch of a new Arm-based PC chip, marking the company's entry into the personal computer market. This significant development was revealed during a recent event and will be featured in upcoming laptops from major manufacturers including Dell, Microsoft, HP, and ASUS. The introduction of this chip aims to enhance performance and efficiency in PCs, tapping into the growing demand for advanced computing solutions. By leveraging Arm architecture, Nvidia seeks to provide a competitive alternative in the rapidly evolving technology landscape, appealing to both consumers and businesses looking for innovative computing options.
Powering the next era of AI in manufacturing: Why it’s time to upgrade to the NVIDIA RTX PRO 4500 Blackwell Workstation Edition
NVIDIA has unveiled its latest advancements in manufacturing technology, showcasing how artificial intelligence and digital twins can significantly accelerate innovation in the industry. This announcement was made during a recent event held in October 2023, where industry leaders gathered to explore cutting-edge solutions. The integration of AI and digital twin technology aims to enhance efficiency and streamline processes within manufacturing operations. By leveraging NVIDIA's Blackwell architecture, companies can expect improved data processing capabilities that facilitate real-time decision-making and predictive analytics. This transformation is poised to not only boost productivity but also drive competitive advantage in an increasingly fast-paced market.
NVIDIA Launches Cosmos 3, the Open Frontier Foundation Model for Physical AI
NVIDIA has unveiled its latest innovation, the NVIDIA Cosmos™ 3, a groundbreaking open world foundation model designed for physical AI. This advanced system integrates a mixture-of-transformers architecture that seamlessly combines vision reasoning, world generation, and action prediction into one cohesive platform. The launch, which took place today, marks a significant step forward in the development of artificial intelligence, aiming to enhance the capabilities of AI in understanding and interacting with the physical world. By leveraging this sophisticated technology, NVIDIA seeks to push the boundaries of AI applications across various industries, paving the way for more intelligent and responsive systems.
Closing the Insight-to-Action Gap: An Integration Architecture for Automated Predictive Maintenance
Recent research highlights that the primary challenge in industrial predictive maintenance is not the inaccuracy of models, but rather the ineffective transition from anomaly detection to actionable response. The study proposes a new integration architecture designed to link machine learning-based anomaly detection systems directly with maintenance execution systems within plants. This innovative approach aims to transform traditional monitoring dashboards into dynamic systems that not only identify issues but also facilitate immediate corrective actions. By addressing the critical gap between detection and response, this integration seeks to enhance operational efficiency and reduce downtime in industrial settings.
Factory / Plant Maintenance
Automate 2026 Q&A with maxon
At a recent industry conference, engineers emphasized the importance of engaging in technical discussions rather than merely showcasing products. The event, held in October 2023, focused on critical topics such as managing duty cycles, understanding thermal limits, and addressing the challenges of torque versus size constraints in engineering design. Participants highlighted the need for collaborative dialogue to evaluate various control architectures, aiming to enhance innovation and efficiency within the field. This approach reflects a growing recognition that deeper engineering conversations can lead to more effective solutions and advancements in technology.
28% of Japanese express concerns about the safety of physical AI, compared to 11% globally.
BlackBerry Limited's QNX division has released a comprehensive report titled "Inside the Robot: An Investigation into Robot Architecture," which surveyed 1,000 robotics engineers. This initiative aims to provide insights into the current state of robotic architecture, reflecting the growing importance of robotics in various industries. The report highlights trends, challenges, and advancements in the field, underscoring QNX's commitment to enhancing the development and integration of robotic technologies. By gathering input from a diverse group of engineers, the study seeks to inform stakeholders about the evolving landscape of robotics and its implications for future innovations.
Navy tests MUSV autonomous control and payload architecture across seven prototypes
The US Navy has approved seven submissions for medium unmanned surface vessels (MUSVs) as part of its ongoing efforts to enhance maritime capabilities. This decision, announced recently, marks a significant step in the Navy's initiative to integrate advanced unmanned technologies into its fleet. The approvals come in response to the increasing demand for innovative solutions to address modern naval challenges, including surveillance and reconnaissance missions. By incorporating these unmanned vessels, the Navy aims to improve operational efficiency and reduce risks to personnel. The selected submissions will now move forward in the development process, with the Navy collaborating closely with the respective contractors to ensure successful implementation. This initiative underscores the Navy's commitment to leveraging cutting-edge technology to maintain its strategic advantage in maritime operations.
GigaAI Unveils Physical AGI "Dual Pyramid" System, Targeting the Embodied Intelligence Scaling Wall
GigaAI unveiled the world's first Physical AGI "Dual Pyramid" architecture during a launch event on May 20 in Wuhan's Optical Valley. This innovative dual-track framework is designed to address the data and algorithmic bottlenecks that have hindered the advancement of embodied AI, aiming to facilitate true scaling in the field. The introduction of this architecture marks a significant milestone in artificial intelligence development, as it seeks to overcome existing limitations and enhance the capabilities of AI systems.
AI
Rajant Health (RHI) and Chord Robotics Expand Cowbell Platform to Enable Scalable, Multi-Domain Collaborative Autonomy
Rajant Health (RHI) and Chord Robotics have announced an expanded partnership to enhance their Cowbell platform, introducing advanced "Flying Cowbell" capabilities aimed at enabling scalable, multi-domain collaborative autonomy. This collaboration, revealed on May 22, 2026, integrates RHI's Kinetic Mesh® networking with Chord Robotics' TEMPO™ software, facilitating real-time control of mixed fleets operating across air, land, and sea. The "Flying Cowbell" system transforms mobile nodes, including unmanned aerial systems (UAS) and unmanned surface vehicles (USVs), into active participants in a distributed compute and autonomy framework. This architecture allows for distributed workload execution, dynamic cluster formation, and transport-agnostic operations, even in connectivity-constrained environments. The partnership aims to address the need for persistent coverage and dynamic mission adaptation in complex scenarios where traditional infrastructure may be lacking. By leveraging Rajant's InstaMesh® networking capabilities alongside TEMPO, operators can manage heterogeneous fleets effectively, ensuring that each vehicle can make independent decisions while collaborating seamlessly. RHI's CEO, Robert J. Schena, emphasized that the initiative represents a shift from static infrastructure to a mobility-native system, while Chord Robotics' CEO, James Cooney, highlighted the potential for scaling autonomous fleets in challenging environments. Together, the companies are poised to redefine collaborative autonomy in unmanned systems, enhancing operational efficiency and adaptability.
Māori Text-to-Speech Model Spurns Big Tech’s Values
Researchers at the University of Waikato in New Zealand have developed a high-fidelity synthetic voice for te reo Māori, the indigenous language of the country, in response to concerns over the ownership and control of Māori language data by foreign technology companies. Led by associate professor Te Taka Keegan and his former master's student Kingsley Eng, the project was motivated by a desire for "sovereign digital systems" that prioritize Māori ownership of their language resources. The initiative began with the recording of 4.5 hours of data from Ngaringi Katipa, a fluent speaker and language mentor, which was later expanded to 7 hours and 45 minutes. The researchers faced challenges due to the unique linguistic features of te reo Māori, such as vowel length and digraphs, which can alter meanings. They employed a phoneme-based approach to training the text-to-speech model, utilizing open-source tools and testing various neural architectures to achieve an effective AI voice with a word error rate of 6.78 percent. Despite receiving funding from Google, Keegan emphasized that the ownership of the voice model remains a collective responsibility of the Māori community, particularly the tribes affiliated with Katipa. The project aims to empower Māori language speakers and establish a framework for similar initiatives among other indigenous communities globally. Keegan envisions a future where community-owned language models can preserve and promote indigenous knowledge, ensuring that technology serves to empower rather than diminish cultural heritage.
Artificial-intelligence Languages Ai-models
Agentic AI for Robot Teams
Researchers at the Johns Hopkins Applied Physics Laboratory are making strides in the development of agentic artificial intelligence aimed at enhancing collaborative robotic teams. During a recent presentation, they outlined the significant challenges associated with achieving autonomy, coordination, and adaptability among diverse robotic systems. To address these issues, the team introduced a scalable architecture designed to facilitate agentic behaviors in multi-robot environments. The presentation also featured demonstrations of this innovative approach, showcasing its application in hardware with a varied group of robots. Additionally, the researchers shared valuable insights gained from their ongoing research and development efforts, highlighting key challenges faced and lessons learned throughout the process. This work not only advances the field of robotics but also sets the stage for future developments in agentic AI technology.
Type-webinar Agentic-ai Robotics Llms
BANNER ENGINEERING - RSio Remote Safe I/O Blocks
Banner Engineering has introduced a new line of I/O blocks designed to enhance safety architectures in industrial settings. These innovative blocks support EtherNet/IP and CIP Safety™, allowing for configurable inputs for emergency stops, light curtains, and switches. The introduction of in-series diagnostics significantly reduces cable usage, enabling the connection of up to 192 safety devices per block while providing device-level status updates. The RSio series is characterized by its scalability and robustness, making it particularly suitable for large conveyor networks and modular machine designs. This development aims to streamline safety management and improve operational efficiency in manufacturing environments.
Perceptron Mk1 shocks with highly performant video analysis AI model 80-90% cheaper than Anthropic, OpenAI & Google
Perceptron Inc., a Bellevue-based startup founded by former Meta researchers Armen Aghajanyan and Akshat Shrivastava, has launched its flagship video analysis model, Mk1, aimed at revolutionizing how enterprises utilize AI in real-time video processing. Announced today, this innovative model is priced significantly lower than competitors, at $0.15 per million input tokens and $1.50 per million output tokens, making it accessible for large-scale industrial applications. The Mk1 model, developed over 16 months, excels in understanding complex physical interactions and temporal reasoning, outperforming established models like OpenAI's GPT-5 and Google's Gemini 3.1 Pro in various benchmarks. Its unique architecture allows it to process video streams continuously, maintaining object identity and providing precise analysis of dynamic scenes, which is particularly beneficial for sectors such as security, robotics, and marketing. Perceptron aims to position Mk1 as a leader in the "Efficiency Frontier," balancing high performance with cost-effectiveness. The model's capabilities extend to auto-clipping highlights from live sports and enhancing quality control in manufacturing. A public demo site is available for potential users to explore its functionalities. This launch signifies a significant step towards integrating advanced AI into real-world applications, as the company seeks to make "physical AI" as prevalent as its digital counterpart.
Technology
Simulation Platforms for Underwater Robotic Applications: Architectures, Capabilities, and Research Directions
A recent study published in the Journal of Field Robotics highlights advancements in autonomous robotic navigation. Researchers from a leading university conducted experiments to improve the efficiency and accuracy of robots in complex environments. The study, released in early October 2023, focused on various terrains, including urban settings and natural landscapes, to assess how robots can better adapt to their surroundings. The motivation behind this research stems from the increasing demand for autonomous systems in industries such as agriculture, logistics, and disaster response. By enhancing the navigation capabilities of robots, the researchers aim to facilitate their deployment in real-world applications, ultimately improving operational efficiency and safety. The team utilized a combination of machine learning algorithms and sensor technologies to develop a new navigation framework. This innovative approach allows robots to process environmental data in real-time, enabling them to make informed decisions and navigate obstacles more effectively. The findings suggest that these advancements could significantly reduce the time and resources required for robots to complete tasks in unpredictable environments. As the field of robotics continues to evolve, this research represents a crucial step towards more reliable and versatile autonomous systems, paving the way for broader applications in various sectors.
SURVEY ARTICLE
Tesla Optimus AGI & End-to-End Neural Nets: The Brain Behind the Robot
Tesla is advancing its artificial intelligence capabilities with the introduction of its latest technologies, including VLA models, the AI5 chip, Grok integration, and Cortex 2.0. These innovations are part of the company's roadmap toward achieving artificial general intelligence (AGI) by 2026. The developments were detailed in a comprehensive technical breakdown that outlines how Tesla's robot AI architecture is designed to enhance performance and efficiency. By leveraging these cutting-edge technologies, Tesla aims to position itself at the forefront of the AI revolution, addressing the growing demand for intelligent automation in various sectors. The insights shared reflect the company's commitment to pushing the boundaries of AI and robotics, with a focus on creating systems that can learn and adapt in real-time. As Tesla continues to refine its approach, the implications of these advancements could significantly impact the future of robotics and AI integration across industries.
ASELSAN at SAHA 2026: Introducing next-generation multi-domain defense systems
At the SAHA 2026 exhibition, ASELSAN introduced its latest defense portfolio, showcasing a comprehensive and integrated defense architecture. This innovative framework combines electronic warfare, counter-unmanned aerial vehicle (UAV) systems, and capabilities for both airborne and naval operations. The unveiling highlights ASELSAN's commitment to advancing military technology and enhancing operational efficiency in response to evolving security challenges. By integrating these diverse capabilities, the company aims to provide a unified solution that meets the complex demands of modern defense environments.
Global Naval Warfare Sponsored Post Air Force Army Aselsan
InfluxData Partners with Litmus to Connect, Contextualize and Store Operations Data
Manufacturers are set to benefit from a new integration that allows for the creation of a unified architecture across edge, on-premises, and cloud deployments. This development, announced recently, aims to streamline operations and enhance efficiency in production processes. By consolidating various deployment environments, companies can improve data management and accessibility, ultimately leading to better decision-making and increased productivity. The integration is expected to be implemented in the coming months, providing manufacturers with the tools necessary to adapt to the evolving technological landscape. This initiative reflects the industry's growing emphasis on digital transformation and the need for cohesive systems that can operate seamlessly across different platforms.
Factory / Analytics
Robotically assembled building blocks could make construction more efficient and sustainable
Recent research indicates that the construction of buildings using interlocking subunits is not only mechanically viable but also significantly reduces carbon emissions. This innovative approach to building design aims to address environmental concerns associated with traditional construction methods. The findings, which emerged from a study conducted by a team of engineers and architects, highlight the potential for a more sustainable future in the construction industry. By utilizing modular components that fit together seamlessly, this method could streamline the building process while minimizing waste and energy consumption. The research, completed in late 2023, emphasizes the urgent need for eco-friendly practices in architecture and construction, as the industry seeks to lower its carbon footprint and contribute to global sustainability efforts.
ABB Robotics launches PickMaster Lite to simplify & accelerate robotic picking
ABB Robotics has introduced PickMaster® Lite, a simplified version of its robotic picking software, aimed at packaging OEMs and system integrators. Launched on May 5, 2026, this new software is designed to accelerate the development of high-speed, vision-guided robotic picking solutions. By offering essential features for common picking tasks, PickMaster Lite reduces engineering efforts by 30% and commissioning time by 25%, while ensuring reliable performance. The motivation behind this launch stems from the increasing demand for automation in manufacturing, driven by labor shortages and consumer expectations for personalized products. Craig McDonnell, Business Line Managing Director at ABB Robotics, emphasized the need for quick and reliable automation solutions to enhance production flexibility. PickMaster Lite employs an intuitive, task-based interface with pre-configured templates, eliminating the need for specialized programming skills. It integrates seamlessly with existing machine control architectures, allowing for easy communication with PLC and HMI systems. This capability enables machine builders to manage key functions directly through their preferred control systems, thus minimizing development risks. The software is particularly suited for high-volume, cost-sensitive applications in sectors such as consumer goods, food and beverage, pharmaceuticals, electronics, and e-commerce. As part of the broader PickMaster family, it offers a scalable solution that can evolve alongside production needs, with options for more advanced functionalities through PickMaster and PickMaster Twin. For additional details, interested parties can visit ABB's robotics website.
Ouster Releases Family of ‘Native Color Lidar’ Sensors for Robotics, Autonomous Vehicles
Ouster has launched a groundbreaking series of digital lidar sensors, the Rev8 family, which the company claims to be the world's first native color lidar platform. This innovative technology is designed for applications in robotics, autonomous vehicles, and industrial AI systems. The announcement was made recently, showcasing the sensors' capabilities powered by Ouster's advanced L4 Silicon architecture, which enhances their range and performance. This development marks a significant advancement in lidar technology, aiming to meet the growing demand for high-quality sensing solutions in various industries.
AI AI Use Cases Robotics autonomous driving industrial AI LiDAR
The AI Server Challenge: Testing Power at Scale
Recent advancements in artificial intelligence (AI) are driving the need for specialized power test systems tailored for next-generation AI architectures. As the demand for faster GPUs and more efficient accelerators grows, the industry recognizes that traditional power testing methods may not suffice. This shift is particularly relevant as AI applications become increasingly complex and resource-intensive, necessitating a reevaluation of existing testing frameworks. The urgency for these purpose-built systems arises from the need to ensure that AI technologies can operate effectively and sustainably. With AI's rapid evolution, companies are seeking innovative solutions to optimize performance while managing energy consumption. The integration of advanced power testing will enable developers to better assess the efficiency and reliability of their AI systems, ultimately leading to more robust and scalable technologies. As the AI landscape continues to evolve, industry leaders are collaborating to design and implement these specialized power test systems, ensuring that they meet the unique demands of next-gen AI workloads. This proactive approach aims to enhance the overall performance and sustainability of AI solutions, paving the way for future breakthroughs in the field.
Closing the latency gap: Why physical AI requires edge-first architectures
Madhu Gaganam, the founder and CEO of Cogniedge.ai, emphasized the necessity for the robotics industry to evolve beyond traditional safety measures such as cages and reduced speeds in response to the growing demand for collaborative robots, or cobots. In a recent statement, Gaganam highlighted that the shift towards true cobots requires innovative approaches, particularly the implementation of edge-first architectures to effectively close the latency gap in physical AI applications. This transition is crucial for enhancing the performance and safety of robots working alongside humans. The insights were shared in a piece featured on The Robot Report, underscoring the importance of adapting technology to meet the changing needs of the industry.
6-Axis Artificial Intelligence Artificial Intelligence / Cognition Assembly Collaborative Robots Design / Development
6-Axis Articulated Robot vs. SCARA Robot: Which is Best for Assembly?
In the evolving landscape of industrial automation, the choice of mechanical architecture is crucial for optimizing production lines. Key players in this field are exploring two primary configurations: SCARA (Selective Compliance Assembly Robot Arm) and articulated robots, alongside the emerging collaborative robots that offer enhanced flexibility and safe interaction with human workers. The SCARA robot, designed for high-speed, linear assembly tasks, excels in pick-and-place and packaging operations but lacks the flexibility to handle complex movements. Conversely, the 6-axis articulated robot mimics human joint movements, enabling it to perform intricate tasks such as inserting screws at angles and navigating tight spaces, making it essential for complex assembly processes. As factories increasingly shift towards high-mix production, the demand for collaborative robots has surged. These systems combine the agility of articulated robots with the safety of human interaction, allowing for complex movements without compromising worker safety. JAKA, a leader in automation solutions, emphasizes the importance of adaptability in modern assembly. Their JAKA A series robots offer the precision of traditional articulated systems while ensuring ease of use and safety. With a repeatability of ±0.02mm, these robots are suited for high-speed assembly and testing. For larger applications, the JAKA Zu series provides diverse payload options, catering to various assembly needs. JAKA's collaborative robots come equipped with an intuitive wireless teaching system, enabling teams to program complex paths quickly, thus enhancing efficiency and flexibility in smart manufacturing.
AI² Robotics Founder Defends VLA, Launches NeuroVLA Model
Guo Yandong, the founder of AI² Robotics, recently defended the VLA architecture as essential for advancing embodied intelligence. During a presentation, he introduced the NeuroVLA, a brain-inspired model designed to enhance cognitive processing in robotics. Additionally, Yandong unveiled the AlphaBrain Platform, an open-source toolkit that supports plug-and-play world models, allowing developers to easily integrate and customize their robotic systems. This initiative aims to foster innovation in the field of robotics by providing accessible resources for researchers and developers. The announcement highlights AI² Robotics' commitment to pushing the boundaries of artificial intelligence and robotics technology.
News
ShengShu Technology Unveils Motubrain: A Unified "World Action Model" to Solve the Robotics Scaling Problem
ShengShu Technology has unveiled Motubrain, an innovative robotic brain designed to integrate perception and action seamlessly. This new technology aims to surpass traditional Variable-Length Architecture (VLA) models in performance on global benchmarks. The announcement was made recently, showcasing ShengShu's commitment to advancing robotics and artificial intelligence. By creating a hardware-agnostic solution, Motubrain allows for greater flexibility and efficiency in robotic applications, potentially transforming various industries that rely on automation and intelligent systems. The development of Motubrain reflects the growing demand for more sophisticated robotic technologies that can adapt to diverse environments and tasks.
WAM China ShengShu Technology world-model MotuBrain
All3 raises $25m in Seed funding to triple productivity in the construction industry through robotics and AI
All3, a London-based startup focused on revolutionizing the construction industry, has successfully raised $25 million in seed funding to enhance productivity through robotics and artificial intelligence. The funding round, led by RTP Global and supported by SuperSeed, Begin Capital, s16vc, and VNV Global, aims to develop an integrated system that includes an AI architecture platform, robotic factories, and Mantis, an autonomous robot designed for on-site assembly. With Mantis already operational and initial commercial deployments set for Germany later this year, All3's innovative approach promises significant cost savings and efficiency improvements, potentially reducing construction timelines by up to 50% and embodied carbon by 25%. The founding team, which previously established the successful grocery delivery service Samokat, is applying their expertise to address the stagnation in construction productivity, which has seen little advancement in the past 50 years. The funding will primarily support research and development efforts in London and Belgrade, as well as the deployment of robots across active construction sites in Germany, where there is a pressing need for approximately 700,000 new homes. CEO Rodion Shishkov emphasized the company's mission to tackle the housing crisis and improve access to quality housing through advanced technology. Early market demand has already validated All3's model, with over 100,000 square meters of residential projects processed using their AI-powered design software, laying the groundwork for a robust construction pipeline in the coming years.
Better Hardware Could Turn Zeros into AI Heroes
Researchers at Stanford University have developed a groundbreaking hardware accelerator named Onyx, designed to enhance the efficiency of artificial intelligence (AI) computations by leveraging the concept of sparsity. This innovation comes in response to the growing energy demands and carbon footprint associated with increasingly large language models (LLMs), such as Meta's recent Llama release, which boasts 2 trillion parameters. Onyx aims to address the limitations of current hardware, which often fails to fully utilize the sparse nature of AI models, where many parameters are effectively zero. By re-engineering the architecture to support both sparse and dense computations, Onyx achieves significant energy savings—consuming up to one-seventieth the energy of traditional CPUs and performing computations eight times faster on average. The development of Onyx reflects a broader trend in AI research, where experts are exploring new algorithms and hardware solutions to mitigate the environmental impact of AI technologies. The team at Stanford plans to expand Onyx's capabilities to support a wider range of AI operations, potentially revolutionizing the field and paving the way for more sustainable AI practices. As the demand for efficient AI solutions grows, Onyx represents a promising step toward balancing performance and energy consumption in machine learning.
Ai-models Gpus Energy-efficiency Data-compressionRobotics needs a service framework.
RSF defines a common language for robot service capability, lifecycle operations, certification pathways, and service-provider networks.
