German robotics startup Microagi has hired Animesh Garg, a Georgia Tech professor and former Nvidia researcher, as chief research officer. Garg is bringing seven members of his research group through Microagi's acquisition of their stealth startup. Financial terms were not disclosed. Microagi raised $55 million this summer in Germany's largest seed round. The company trains robots for factory jobs in automotive, logistics, and food industries using data collected from workers. The move reflects a broader trend of robotics researchers joining startups. Carnegie Mellon professors cofounded Skild AI, which raised over $2 billion. Stanford's Chelsea Finn and UC Berkeley's Sergey Levine cofounded Physical Intelligence, raising over $1 billion. Garg cited robotics' shift from lab demonstrations toward commercial deployment as his reason for joining.

Nvidia's influence on robotics startups and emerging AI innovations. Published By: Antonio Robles Robles.AI Build AI That Earns Trust. AI Assistant: Mike Electrum Engineering a smarter tomorrow. September 28, 2026 at 06:00 PM PDT robotics AI technology Table of contents. Nvidia's robotics lab and its startup ecosystem. Nvidia, a giant in the chip industry, has been instrumental in laying the groundwork for physical AI, a term popularized by its CEO Jensen Huang. This has led to a wave of new robotics startups founded by former Nvidia employees. Notable among them is Amulya Vishwanath, who left Nvidia to start Techable Ventures, focusing on physical AI. Sanja Fidler, a former vice president of AI research at Nvidia, also founded Veeda AI, which creates simulated worlds for training robots. Nvidia's strategy revolves around developing the chips and software that power these robots, creating a fertile ecosystem for new companies. Nvidia's approach involves building the foundational technology that enables robots to interact with the real world. This includes the development of AI chips and training software. The company's ecosystem supports various startups, such as Flexion Robotics and Dyna Robotics, by providing them with early access to tools and biweekly meetings for feedback and development. According to Nikita Rudin, CEO of Flexion, "We were the beta testers - and now again, just from the outside." The implications of Nvidia's strategy are significant for the robotics industry. With physical AI companies raising $33.4 billion globally in the first half of 2026, surpassing last year's total of $28.7 billion, the market is ripe for innovation. Nvidia's alumni are targeting diverse areas, from warehouse robotics to physical task automation, creating a cycle where more startups lead to increased demand for Nvidia's infrastructure. From my perspective, Nvidia's influence in the robotics sector exemplifies a strategic alignment of hardware and software development, fostering an environment where innovation thrives. This trend aligns with broader industry movements towards more specialized and capable robotic systems, as seen in the increasing venture funding and real-world applications of these technologies. Microagi's recruitment of university researchers. Microagi, a German robotics startup, has recently made headlines by hiring Animesh Garg, a former Nvidia researcher and Georgia Tech professor, as its chief research officer. Garg, along with seven members of his research group, joined Microagi following the acquisition of their stealth startup. This move underscores the growing trend of university researchers transitioning to the commercial sector, driven by the influx of billions into physical AI. Microagi's focus is on adapting general-purpose robotics models for specific tasks, using data collected from workers to tailor these models to different industries such as automotive and logistics. Garg's expertise in simulation technology, previously developed at Nvidia, aligns with Microagi's approach of using human demonstration data to train robots. The startup's recent $55 million seed round, the largest for a German startup, highlights the financial backing and confidence in its innovative methodologies. The recruitment of academic talent like Garg signifies a shift towards large-scale commercial deployment of robotics technologies. As Garg stated, "We have gotten to a point where we can create sustainable businesses." This transition from lab demonstrations to real-world applications is fueled by advances in AI, reduced hardware costs, and the need for automation in manufacturing. In the context of Robles.AI, the integration of academic research with commercial applications presents opportunities for collaboration and innovation. By leveraging its expertise in AI and robotics, Robles.AI can contribute to the development of adaptive systems that meet the evolving needs of various industries. The rise of Jev: A new AI model from TypeSafe AI. TypeSafe AI has introduced Jev, a novel AI model that is gaining traction in San Francisco's tech scene. Unlike traditional large language models (LLMs), Jev focuses on decision-making rather than text generation. This model is characterized by its speed and cost-effectiveness, making it a compelling alternative to existing AI solutions. Jev operates by rapidly responding to structured questions with probability estimates of accuracy. It excels in tasks such as classifying customer support queries and moderating online chat feeds. Allie Laabs, TypeSafe's head of developer relations, explained that Jev's strength lies in breaking down questions into specific components, allowing for quick and precise responses. This approach challenges the dominance of LLMs by offering a new paradigm in AI development. The implications of Jev's introduction are significant for the AI industry. With venture capitalists valuing TypeSafe at $10 billion, the model's potential to revolutionize AI applications is evident. Developers have already integrated Jev into various products, enhancing speed and reducing costs. As Lawrence Liang, a developer, noted, Jev's integration into his personal AI agent improved its efficiency by 80%. From a broader perspective, Jev represents a shift towards more specialized AI models that prioritize decision-making capabilities. This aligns with trends in the industry towards creating AI systems that can operate in dynamic environments, a focus that resonates with Robles.AI's mission to develop intelligent solutions for complex challenges. AMD's acquisition of World Labs. AMD has announced its acquisition of World Labs, an AI model startup led by Fei-Fei Li, in an all-stock deal valued at $8.2 billion. This acquisition is part of AMD's strategy to enhance its AI capabilities and integrate World Labs' expertise in spatial intelligence models. Fei-Fei Li will join AMD as executive vice president and chief scientist, continuing her work on AI model research. World Labs specializes in creating interactive 3D environments from text, images, and video, with applications in robot learning and simulation. Their product, Marble, generates 3D worlds, while the Atlas model predicts new camera views from 2D images. These technologies align with AMD's focus on developing compute platforms for next-generation AI applications. The acquisition of World Labs by AMD highlights the growing importance of AI in shaping future technology roadmaps. As Lisa Su, CEO of AMD, stated, "Building the compute platforms for the next generation of AI requires a deep understanding of how models are evolving." This deal positions AMD to capitalize on the increasing demand for AI-driven solutions in various industries. For Robles.AI, the integration of spatial intelligence models into its offerings can enhance its capabilities in developing adaptive systems for real-world applications. By collaborating with industry leaders like AMD, Robles.AI can leverage cutting-edge technologies to deliver innovative solutions to its clients. Where Robles.AI fits. Robles.AI is uniquely positioned to capitalize on the advancements discussed in these articles. Its focus on developing intelligent robotics solutions aligns with the trends highlighted by Nvidia's influence in the startup ecosystem and Microagi's recruitment of academic talent. By integrating decision-making models like Jev into its systems, Robles.AI can enhance the efficiency and adaptability of its AI solutions. Furthermore, the acquisition of World Labs by AMD presents opportunities for collaboration in the development of spatial intelligence models. By leveraging these technologies, Robles.AI can continue to innovate and provide cutting-edge solutions that meet the evolving needs of its clients in various industries. Its commitment to advancing AI and robotics technologies ensures that Robles.AI remain at the forefront of industry developments, delivering value through innovative and adaptive solutions. As Robles.AI continue to explore new opportunities and partnerships, Robles.AI is poised to lead the way in the next generation of intelligent systems. #Nvidia #physicalAI #Microagi #Jev #WorldLabs Sources. Tokens used: Prompt 6687, Completion 3900, Total 10587

microSLAM: Turning plain monocular RGB videos into interactive RL environments. By microagi research · September 10, 2026 The goal of microagi is to deploy billions of robots worldwide. Western industry is losing its competitiveness. It is held back by a critical shortage of workers for millions of monotonous and dangerous manual jobs, the jobs that form the backbone of its real economy. At microagi, MicroAGI, Inc. is bringing the value of foundation models to the physical world by building the robotics data and deployment infrastructure, right here in the heart of Europe. Shift records how physical work is performed and turns it into training-ready data without teleoperation or interruption to ongoing operations. Atlas uses that data to train models for specific tasks and carries the right robotic system through deployment and reliable production operation. Together, they create a continuous path from human expertise to deployed robots, and from every deployment back into better data and models. The deployment advantage The models will be ready for everyone at the same time. The data won't. * 01Project scoping Without MicroAGI, Inc. * 01No data collected Announcing microSLAM. Today MicroAGI, Inc. is announcing microSLAM, the new state of the art in monocular SLAM from human egocentric video, built by its Zürich-based computer vision team in collaboration with the Computer Vision and Geometry Group at ETH Zürich. microSLAM ranks first among monocular SLAM systems on the LaMAria benchmark. Among all systems evaluated on the benchmark, including methods that use additional cameras and inertial measurement units (IMUs), it ranks second overall. At the category level, microSLAM achieves the highest score of any evaluated system on two challenges that are particularly important for robotics: moving-platform sequences and long sequences. It scores 85.6 on moving platform, 30.6 points above Meta's Aria SLAM baseline at 55.0, and 82.3 on long sequences, 11.4 points above the Aria baseline at 70.9. microSLAM achieves these results from monocular input, while Aria SLAM uses binocular cameras and an inertial measurement unit (IMU). #1 Monocular SLAM on the LaMAria benchmark #2 Across all SLAM systems on the LaMAria benchmark From plain video to interactive RL environments. microSLAM turns plain monocular RGB video into fully interactive 3D environments. These are so-called reinforcement learning (RL) environments. They teach state-of-the-art behavioral foundation models to understand their surroundings, navigate across factory floors, and become experts at the industrial use cases its clients care about. Why existing SLAM systems didn't meet its requirements for real robotic deployments. MicroAGI, Inc. tried the existing SLAM systems, and they failed for robotics because of three issues. * The static-world assumption does not hold. Egocentric video is full of motion: hands, people, and objects moving through the scene. * Scale is ambiguous. A single camera cannot tell the true scale of what it sees. * Long sequences must stay consistent. Many productive sequences in robotics span tens of minutes, and potentially hours. ORB-SLAM3, one of the most commonly used SLAM systems, suffers from exactly these three issues. When its estimated poses are fused into a dense reconstruction, accumulated inconsistency remains visible as geometric distortion and misalignment. For the qualitative comparisons in this article, MicroAGI, Inc. keep the camera poses estimated by each SLAM system fixed and fuse the same dense depth source into a truncated signed distance field (TSDF). MicroAGI, Inc. back-project each depth map using the corresponding estimated camera pose, accumulate the observations in the TSDF volume, and extract the final mesh. Because the estimated poses are not altered, pose drift and inconsistency remain visible in the resulting reconstruction. How MicroAGI, Inc. designed microSLAM. MicroAGI, Inc. tackled these challenges in two key ways. * Dynamic-object-aware recognition microSLAM identifies moving objects in the scene and keeps the static-world assumption intact. * Robust loop closure microSLAM uses rich features from the environment to power one of the most robust loop-closure systems to date. The sales-floor sequence below compares the same scene reconstructed by microSLAM and ORB-SLAM3, making the effect of robust loop closure directly visible. microSLAM is one of the foundational steps bringing MicroAGI, Inc. closer to its goal of deploying billions of robots worldwide. MicroAGI, Inc. will release the microSLAM paper and open-source code in the near future, providing the community with a reproducible foundation to evaluate, extend, and build on this work.