Warthog Mobile Manipulator Featuring Large UR20 Arm

Apple suppliers race to Hong Kong: $4bn in a week

Two of Apple's largest Chinese suppliers, Luxshare Precision and Lingyi iTech, are racing to list on the Hong Kong stock exchange in the same week, together raising roughly $4bn. Both companies are using the capital to pivot from smartphone component manufacturing toward AI hardware and humanoid robotics. Lingyi raised $1.1bn at the top of its price range and is targeting 500,000 humanoid robots annually by 2030, while Luxshare is gauging interest for a ~$3bn listing. The rush is partly driven by a June filing deadline that lets companies avoid refiling financials. Hong Kong IPO proceeds are on track for a six-year high of $43bn in 2026. Analysts note caution: Hong Kong shares priced at steep discounts to mainland listings, China's humanoid robot sector is crowded with over 150 competitors, and demand at scale remains unproven.

#RoboCup2026 – humanoid league day 2

Day 2 of RoboCup 2026's humanoid league features 17 countries and teams across three divisions (small, middle, large). CAU Mountain&Sea leads the small division with a perfect record. B-Human dominates the middle division with a +35 goal difference. Tsinghua Hephaestus is the sole unbeaten team in the large division. Notable robot stats include the heaviest competitor at 48kg/160cm (ALICE 4) and the lightest at 3.8kg/53cm (ITAndroids' Chape). Knockout rounds begin tomorrow, with 12 teams per division advancing.

China approves Unitree’s Shanghai listing registration, valuing IPO at $619M

Cơ quan quản lý chứng khoán Trung Quốc đã chấp thuận đợt IPO của Unitree Robotics trên sàn STAR Market (Thượng Hải) với giá trị định giá khoảng 619 triệu USD (4,2 tỷ nhân dân tệ). Sau khi vượt qua mọi rào cản pháp lý, Unitree vẫn cần ấn định giá cổ phiếu cuối cùng và ngày phát hành.

Những công ty robotics nhân tạo đang cạnh tranh trên thị trường công khai, và Unitree với mô hình kinh doanh hiệu quả cùng giá trị công ty cao là ví dụ điển hình cho những cơ hội và thách thức trong ngành công nghệ mới này.

How Open Models Are Driving AI Research

NVIDIA's open model families — Nemotron, Cosmos, Isaac GR00T, and BioNeMo — are playing a central role at ICML 2026, with 145 papers citing Nemotron and hundreds more drawing on other NVIDIA open models. Key research themes include robot world models (e.g., DreamDojo built on Cosmos), AI for life sciences (protein mutation prediction, drug discovery via KERMT), and synthetic data generation. The ecosystem extends beyond NVIDIA's own labs, with companies like Sakana AI, KiloCode, NAVER, Merck, and multiple robotics firms building on these open foundations. Nemotron is increasingly used as a full research stack — open weights, datasets, and recipes — rather than a single model release.

Weave Robotics Isaac 1: a $7,999 home robot

Weave Robotics has unveiled Isaac 1, a $7,999 wheeled home robot that handles laundry-related chores like folding clothes, making beds, and tidying shoes. Unlike humanoid rivals such as 1X's Neo (~$20,000) or Tesla's Optimus, Isaac 1 uses a wheeled base and claw grippers instead of legs and fingers, deliberately limiting its scope to reduce cost. The Y Combinator-backed startup offers it at $449/month as an alternative. Deliveries begin in September 2026 in California only, with the rest of the US following in 2027. Partial autonomy is supplemented by remote teleoperation, and privacy concerns exist around whether home footage is used for training data.

LeRobot v0.6.0: Imagine, Evaluate, Improve

LeRobot v0.6.0 is a major release for the open-source robotics framework, introducing world model policies (VLA-JEPA, FastWAM, LingBot-VA) that learn to predict the future during training at zero extra inference cost. The release adds five new VLAs including GR00T N1.7, MolmoAct2, EO-1, EVO1, and Multitask DiT. A new unified reward models API ships with Robometer (a 4B general-purpose reward model trained on 1M+ trajectories) and TOPReward (zero-shot VLM-based reward). Six new simulation benchmarks are unified under the lerobot-eval CLI. Training improvements include FSDP multi-GPU support, cloud training via HF Jobs, and up to 2x faster data loading. Datasets gain depth camera support, automatic language annotation via VLMs, and flexible video encoding. A new lerobot-rollout CLI enables DAgger-style human-in-the-loop corrections for iterative policy improvement.

#RoboCup2026 – humanoid league knockout stages

RoboCup 2026 in Incheon, South Korea concluded its humanoid league competitions, with winners crowned across small, middle, and large divisions. In the small division, Invic from Wuhan University took first place. B-Human from Universität Bremen and DFKI won the middle division, while Tsinghua Hephaestus claimed the large division title. Five league-wide awards were also presented, including Best Humanoid Software to B-Human and Best Innovation to Bahia Robotics Team. The event continues with a symposium on robotics and AI research.