Embodied Intelligence Breakthrough: Amap Fully Open-Sources Universal Robot Base Model ABot-M0

marsbitPublicado em 2026-04-01Última atualização em 2026-04-01

Resumo

Embodied Intelligence Breakthrough: AutoNavi Open-Sources Universal Robot Base Model ABot-M0 AutoNavi has announced the full open-source release of ABot-M0, the world's first unified architecture-based embodied manipulation base model. This model is designed to enable "one general brain to adapt to multiple forms of robots," aiming to break down barriers between heterogeneous hardware and accelerate the adoption of embodied intelligence in industrial and household settings. ABot-M0 demonstrated exceptional performance in industry tests, achieving a task success rate of 80.5% on the Libero-Plus benchmark—a nearly 30% improvement over the previous benchmark, Pi0. It also set new state-of-the-art records on benchmarks like Libero and RoboCasa. The open-source release addresses long-standing challenges in the field, such as data isolation and deployment difficulties, by providing resources across three key dimensions: - **Data:** The UniACT dataset, the largest of its kind, with over 6 million real operation trajectories and full data pipeline tools. - **Algorithm:** The model architecture and training framework, featuring innovative components like Action Manifold Learning (AML) and a dual-stream perception architecture. - **Model:** End-to-end pre-trained models and a complete toolchain for out-of-the-box deployment, significantly lowering the barrier to adaptation. According to AutoNavi's ABot-M0 technical lead, this open-source initiative aims to build a bridge between ac...

The field of embodied intelligence has reached a milestone. Amap today officially announced the full open-source release of the world's first unified architecture-based robot operation base model ABot-M0. The core positioning of this model is to achieve "one universal brain adaptable to various forms of robots," aiming to break down barriers between heterogeneous hardware and accelerate the transition of embodied intelligence from the laboratory to industrial and home scenarios.

Core Technology and Performance

ABot-M0 has demonstrated outstanding performance in multiple industry authoritative benchmark tests. Data shows that the model achieved a task success rate of up to 80.5% on the Libero-Plus benchmark, a nearly 30% improvement over the previous industry benchmark solution Pi0. Furthermore, it set new SOTA (State-of-the-Art) records in tests such as Libero and RoboCasa.

Full Open-Source Across Three Dimensions

To address the long-standing pain points of "data silos" and "deployment difficulties" in the field of embodied intelligence, Amap's open-source release covers three key dimensions: underlying data, core algorithms, and pre-trained models:

  • Data Level: Open-sourced the currently largest universal robot dataset UniACT. This dataset integrates over 6 million real operation trajectories and provides a complete processing pipeline from heterogeneous data to standardized training data.

  • Algorithm Level: Simultaneously released the model architecture and training framework. Core highlights include Amap's innovative Action Manifold Learning (AML) algorithm and Dual-Stream Perception Architecture, endowing robots with exceptional spatial understanding and action execution capabilities.

  • Model Level: Provided end-to-end pre-trained models and a complete toolchain. Developers can achieve "out-of-the-box" usability without building a framework from scratch, significantly lowering the barrier to adapting to industrial collaborative or home service robots.

Industry Impact

Amap's ABot-M0 technical lead stated that true general embodied intelligence requires the collective refinement of global developers. The open-sourcing of ABot-M0 is not just a sharing of technology but also aims to build a bridge connecting academic research and industrial application, enabling every robot of different forms to possess a smart, reliable, and universal "brain".

Perguntas relacionadas

QWhat is the name of the general-purpose robotic base model that AutoNavi has fully open-sourced?

AThe model is called ABot-M0.

QWhat is the core positioning or main goal of the ABot-M0 model?

AIts core positioning is to achieve 'one general-purpose brain adapted to various forms of robots', aiming to break down barriers between heterogeneous hardware.

QWhat is the name of the large-scale dataset that was open-sourced alongside the model, and how many real operation trajectories does it contain?

AThe dataset is called UniACT, and it integrates over 6 million real operation trajectories.

QOn which benchmark did ABot-M0 achieve a task success rate of 80.5%, and what was the performance improvement over the previous benchmark solution?

AIt achieved an 80.5% success rate on the Libero-Plus benchmark, which is a nearly 30% improvement over the previous benchmark solution, Pi0.

QName two core technical highlights of the algorithm that were open-sourced.

AThe two core technical highlights are the Action Manifold Learning (AML) algorithm and the Dual-Stream Perception architecture.

Leituras Relacionadas

Platform Token Surges 15-Fold in Half a Month, Pons Tops Both Token Issuance and Trading on Robinhood Chain

The article details the rapid rise of Pons as the leading token launch and trading platform on Robinhood Chain. Following NOXA's halt of new token issuance, several platforms vied for the resulting traffic. By mid-July, Pons surged ahead, achieving the highest daily token creation volume (over 15,000 tokens on July 15th) and soon after capturing the largest daily trading volume ($40.7M on July 18th), consistently maintaining both top positions. Pons's key advantage is its streamlined, single-transaction process that deploys a token's contract and its Uniswap V3 liquidity pool simultaneously, eliminating the need for bonding curves or subsequent DEX migrations. This simplicity caters to the demand for fast, low-cost launches on Robinhood Chain. The platform's native token, PONS, played a pivotal role in its growth, experiencing a price surge of over 15x from mid-July. A tokenomics model that uses 80% of protocol fees to buy back and burn PONS has contributed to its value and heightened platform visibility. Looking ahead, Pons V2 aims to expand beyond meme tokens by integrating support for paying transaction fees with assets like USDG or RWA tokens, aligning with Robinhood Chain's focus on on-chain finance and real-world assets.

marsbitHá 29m

Platform Token Surges 15-Fold in Half a Month, Pons Tops Both Token Issuance and Trading on Robinhood Chain

marsbitHá 29m

Apple's Q3 Results Largely Meet Expectations, Supply and Memory Pressure Drive After-Hours Drop of 6%

Apple's Q3 (FY2026) earnings largely met expectations, but weak guidance and rising costs drove the stock down 6.3% after hours. Revenue of $109.42B slightly beat estimates, and EPS of $2.02 was boosted by a $0.11 tariff refund. Core EPS was $1.91, aligning with forecasts. iPhone and Mac sales were strong, but iPad and Services revenue fell short. Greater China growth slowed. The outlook for the next quarter is concerning. Apple guided for 9-11% revenue growth, below the ~12% consensus, citing significant supply constraints for iPhones, Macs, and iPads due to advanced chip shortages. Gross margin guidance of 47-48% includes a 1-point tariff benefit, implying a core margin of 46-47%, down from 48.1% this quarter. Management highlighted rising memory costs as a major pressure point. While Apple's AI initiatives like Siri are progressing, monetization remains early. With the stock near all-time highs, the combination of in-line results, segment weaknesses, and a cautious outlook triggered significant profit-taking. The focus has shifted from resilient demand to supply chain execution and cost management.

marsbitHá 34m

Apple's Q3 Results Largely Meet Expectations, Supply and Memory Pressure Drive After-Hours Drop of 6%

marsbitHá 34m

Platform Token Soars 15x in Half a Month, Pons Tops Robinhood Chain Charts for Both Token Launches and Trading

In just half a month, the Pons platform token (PONS) surged over 15x in value as Pons solidified its position as the leading token launch and trading platform on Robinhood Chain, capturing both the highest daily token creation volume and transaction volume. Following the exit of previous leader NOXA, competition initially favored Flap before Pons emerged dominant around mid-July. By July 26th, Pons accounted for 77.1% of total trading volume across all Robinhood Chain launchpads. Pons's success is attributed to its streamlined, one-transaction process for deploying a token contract and its corresponding Uniswap V3 liquidity pool simultaneously, eliminating the need for bonding curves or complex migrations. This simplicity, combined with low creation costs and immediate tradability, attracted significant creation activity and speculative trading. The meteoric rise of its native token, PONS, further fueled platform growth. A tokenomics model directing 80% of protocol fees to buy back and burn PONS created a reinforcing cycle: platform usage boosted PONS value, which in turn attracted more users. Looking ahead, Pons V2 aims to expand beyond meme tokens by integrating support for paying fees in stablecoins (USDG) and RWA assets, aligning with Robinhood Chain's broader focus on on-chain finance and stock tokens.

Odaily星球日报Há 34m

Platform Token Soars 15x in Half a Month, Pons Tops Robinhood Chain Charts for Both Token Launches and Trading

Odaily星球日报Há 34m

Qualcomm Chip Price Hike Deals a Blow to Android Phones

Qualcomm has officially announced a new round of price increases for its entire chip portfolio, effective September 1. The hikes, reaching up to 18% for the flagship Snapdragon 8 Elite Gen 6 Pro, follow similar moves by MediaTek, intensifying cost pressures on the already strained smartphone industry. CEO Cristiano Amon confirmed the plan, citing the need to offset rising industry-wide costs and restore declining profit margins. Qualcomm's Q3 FY2026 results showed a 25% drop in net profit, with mobile revenue plunging 20% year-over-year, hitting its lowest level since 2021. The surge in AI computing demand has led memory manufacturers to prioritize HBM production, creating a shortage in general-purpose DRAM and NAND Flash chips. Their prices soared by 93%-98% and 55%-60% respectively in Q1 2026, causing the memory cost share in smartphones to jump from 10%-15% to over 30%. Coupled with the soaring cost of advanced nodes like TSMC's 2nm and packaging, overall chip costs have reached historic highs. These upstream pressures are forcing downstream smartphone brands like Xiaomi, OPPO, and vivo to cut orders for mid-to-low-end models by up to 20% and use cost-saving measures like older chipsets. Reportedly, the Snapdragon 8E5 will be repurposed as a "long-lasting" chip for sub-brand phones in H2 2026. Amid this cost crisis, the Android market remains sluggish. Q2 2026 smartphone shipments in China fell 4.3% year-over-year, marking five consecutive quarters of decline. Major Android brands saw market share drop, while Huawei and Apple, with their in-house chip advantages, gained share. Qualcomm is diversifying into automotive and IoT sectors to reduce reliance on smartphones, but these new segments cannot yet fill the mobile revenue gap. Industry observers warn that the full impact of component cost hikes will hit in the second half of 2026, likely leading to higher-than-expected price increases for Android flagships and a further contraction in the Chinese smartphone market.

marsbitHá 45m

Qualcomm Chip Price Hike Deals a Blow to Android Phones

marsbitHá 45m

Breaking: GPT-5.6 Prices Slashed Effective Today

OpenAI has announced significant price cuts for its GPT-5.6 model API, effective immediately. The entry-level **GPT-5.6 Luna** sees the most drastic reduction, with input prices dropping 80% to $0.20 per million tokens and output prices falling to $1.20 per million tokens. The mid-tier **GPT-5.6 Terra** is reduced by 20%, now costing $2.00 (input) and $12.00 (output) per million tokens. The flagship **GPT-5.6 Sol** maintains its original price but introduces a new **Fast mode**, offering speeds up to 2.5 times faster for double the cost. The company attributes these price reductions to efficiency gains achieved through **GPT-5.6 Sol's own involvement in optimizing its production systems**. The model assisted in rewriting GPU kernels and improving speculative decoding, leading to a 20% reduction in end-to-end service costs and over 15% improvement in token generation efficiency. OpenAI emphasizes this process remained human-led. A key focus of the降价 is to lower the barrier for running **AI agent workflows**. By making the capable, tool-calling Luna model significantly cheaper, OpenAI aims to enable more frequent use in cost-sensitive, high-volume tasks like code review and monitoring. This creates a potential feedback loop: model-assisted efficiency gains lead to lower costs, which enables broader agent deployment, which in turn drives further optimization. The new pricing and features will also apply to Codex and ChatGPT Work subscriptions. The changes intensify competition in the large language model market, with OpenAI directly challenging rivals like Anthropic to respond.

marsbitHá 1h

Breaking: GPT-5.6 Prices Slashed Effective Today

marsbitHá 1h

Trading

Spot
活动图片

Categorias popularesright-arrow

Etiquetas Popularesright-arrow