Tesla said on Wednesday it has begun producing Optimus humanoid robots at its Fremont, California, factory, a site previously dedicated to electric vehicle manufacturing. “The robots coming off this line are the first mass-production-grade robots we have built,” Tesla wrote in its third-quarter shareholder letter. “While we will continue to improve the hardware and manufacturing process, we have proven that Optimus’ current manufacturing process is suitable for a commercial product.” CEO Elon Musk said on a conference call that he views the output as “early production” and that the ramp will be slow and “somewhat bumpy, because it is an extremely aggressive new product.” He said Tesla aims to produce 20,000 Optimus robots per week by the end of 2027....
Robotics startup Eka Robotics said it has raised $150 million in a Series A round at a post-money valuation of $1.45 billion, co-led by a16z and F-Prime Capital. The company, founded in 2025 by MIT professor Pulkit Agrawal and former Google DeepMind researcher Tuomas Haarnoja, focuses on developing Vision-Force-Action (VFA) models and dexterous robotic grippers. Eka said its VFA model is a robot foundation model that, unlike vision-language-action (VLA) models, places greater emphasis on force feedback when robots contact objects. It combines visual perception, force sensing and motion control so robots can adjust actions in real time based on contact conditions. For training, Eka uses a combination of physics simulation and large-scale reinforcement learning. Simulators model the different forces a...
Unitree Technology has unveiled the H2-X, a humanoid robot the Chinese company is presenting as capable of martial-arts-style movement and emergency response, according to a company video. The footage shows the robot, which Unitree says can run for up to about eight hours, carrying out high-impact maneuvers: denting a car door and front hood, vaulting onto a vehicle roof, landing from a 4-meter platform with a forward roll and recovering after repeated airborne kicks. In staged rescue and protection scenes, the H2-X lifts the front of a car to free a trapped mannequin, dives to push another mannequin away from an oncoming vehicle and intervenes against an attacker threatening a person. Unitree said the combination of long endurance and burst-power...
Magic Atom said on Oct. 19 that it has launched the MagicArm R1, a flexible biomimetic force-controlled robotic arm, and built a desktop robotic-arm solution around a vision-language-action (VLA) embodied AI model. The solution uses the MagicArm R1 as its physical platform and runs on EchoPolicy-0, the company’s embodied AI model, which ranked first on the MolmoSpaces Combined leaderboard with a 75.4% average success rate in robot manipulation. MolmoSpaces is a global zero-shot embodied AI benchmark initiated by the Allen Institute for AI (AI2). Magic Atom said the system integrates visual perception, intelligent decision-making, motion control, data collection, model training and real-robot reinforcement learning into a single hardware-software platform. It is aimed at robot algorithm research, model training, intelligent operation...
Recently, Galileo Robotics officially launched its all-terrain mobile intelligence platform—the Galileo X Embodied Ground Mobility System. Leveraging its proprietary multi-form adaptive switching technology, the system breaks through the scene- and form-related barriers of traditional robots, effectively addressing the pain points of automation implementation in the industry and accelerating the commercialization of embodied intelligence worldwide. For a long time, poor environmental adaptability has been the core bottleneck hindering the widespread adoption of mobile robots globally. Traditional wheeled, tracked, and quadruped robots each have their own shortcomings and cannot simultaneously meet the diverse demands of high-precision indoor transport, long-distance heavy-load transport outdoors, and obstacle navigation across complex terrain. The industry generally requires the procurement of multiple types of robots to work in...
UBTECH Robotics has signed a strategic memorandum of understanding with the Government of Ras Al Khaimah’s Office of International Relations to establish a comprehensive cooperation framework for advanced humanoid robot research and development, real-world testing and system deployment, and to advance development and application of humanoid robot technologies suited to the Gulf region. The signing was attended by Sheikh Saud bin Saqr Al Qasimi, Supreme Council Member and Ruler of Ras Al Khaimah; Mohamed Abdulrahman Al Hawi, Undersecretary of the UAE Ministry of Investment; Ouyang Yongqin, advisor for Greater China affairs at the Ras Al Khaimah Ruler’s Office; Zhang Ju, chief financial officer of UBTECH Robotics; and Chen Xiaoyun, overseas sales director of UBTECH Robotics. Under the plan, UBTECH will...
Singapore-based embodied AI startup Ropedia said it has released Ropedia W0, a cross-embodiment mobile manipulation world model developed with researchers from Nanyang Technological University, Duke University, the University of California San Diego and Tsinghua University. The model uses a unified action representation to predict first-person future scene changes after actions by humans, robotic arms and humanoid robots, allowing one model to cover different robot embodiments. Ropedia said W0 ranked first across five evaluation benchmarks and 11 tests. In the Ψ₀ and ω-HOME benchmarks, W0 outperformed comparison models trained on full data without post-training, the company said. On the DreamDojo official benchmark, W0 exceeded DreamDojo trained on full data using only 1% of post-training data. The model takes initial frames, text...
WorldOracle, which its developers describe as the first benchmark for action-driven world models that uses real-world execution as the yardstick, launched on Oct. 16 and opened evaluation submissions to developers. The project was jointly completed by researchers from AGIBOT, NVIDIA, Princeton University, the University of Texas at Austin and the University of Toronto, combining world-model developers with benchmark researchers. The platform addresses a question the industry has lacked a benchmark to answer directly: given a set of actions, how far is a model’s predicted future from what actually happens? WorldOracle defines this as “Execution Coherence.” The researchers said a reliable world model must withstand three stress tests: causal obedience, meaning it must follow the given action sequence, including wrong or...
SoftBank is considering an acquisition of Allied Universal that would value the security services company’s equity at up to $18 billion, with talks still at an early stage. The potential deal is linked to Masayoshi Son’s robotics investment push. Reuters reported that SoftBank hopes robots will eventually take on some security duties and replace some employees. Allied Universal has about 800,000 employees worldwide and annual revenue of about $23 billion, according to information the company disclosed this year. It operates in more than 100 countries and serves more than 400 Fortune 500 companies across commercial real estate, healthcare, manufacturing, transportation and data centers. In January, SoftBank led a $1.4 billion Series C funding round for Skild AI, valuing the robot-model...
ACE Robotics has released a visual-spatial grounding-enhanced embodied foundation model, ACE-Brain-0.6, and said the 9-billion-parameter system outperformed OpenAI’s GPT-6 Astra on six of eight grounding benchmarks. The company said the model is designed to improve target identification, spatial localization and interaction anchoring—capabilities it described as critical for translating high-level task reasoning into reliable physical action. It builds on the earlier ACE-Brain-0.5 and focuses on visual grounding: identifying target objects, interaction areas and spatial references, and linking them to position, displacement and motion parameters. In seven public benchmarks, ACE-Brain-0.6 averaged 74.53 points versus 73.68 for GPT-6 Astra, according to the company. It said the model beat GPT-6 Astra on six of eight evaluations, including Where2Place (76.55 vs. 71.10), PixMo-Count (89.41 vs....
