MAROKO133 Update ai: ByteDance Introduces Astra: A Dual-Model Architecture for Autonomous

📌 MAROKO133 Breaking ai: ByteDance Introduces Astra: A Dual-Model Architecture for

The increasing integration of robots across various sectors, from industrial manufacturing to daily life, highlights a growing need for advanced navigation systems. However, contemporary robot navigation systems face significant challenges in diverse and complex indoor environments, exposing the limitations of traditional approaches. Addressing the fundamental questions of “Where am I?”, “Where am I going?”, and “How do I get there?”, ByteDance has developed Astra, an innovative dual-model architecture designed to overcome these traditional navigation bottlenecks and enable general-purpose mobile robots.

Traditional navigation systems typically consist of multiple, smaller, and often rule-based modules to handle the core challenges of target localization, self-localization, and path planning. Target localization involves understanding natural language or image cues to pinpoint a destination on a map. Self-localization requires a robot to determine its precise position within a map, especially challenging in repetitive environments like warehouses where traditional methods often rely on artificial landmarks (e.g., QR codes). Path planning further divides into global planning for rough route generation and local planning for real-time obstacle avoidance and reaching intermediate waypoints.

While foundation models have shown promise in integrating smaller models to tackle broader tasks, the optimal number of models and their effective integration for comprehensive navigation remained an open question.

ByteDance’s Astra, detailed in their paper “Astra: Toward General-Purpose Mobile Robots via Hierarchical Multimodal Learning” (website: https://astra-mobility.github.io/), addresses these limitations. Following the System 1/System 2 paradigm, Astra features two primary sub-models: Astra-Global and Astra-Local. Astra-Global handles low-frequency tasks like target and self-localization, while Astra-Local manages high-frequency tasks such as local path planning and odometry estimation. This architecture promises to revolutionize how robots navigate complex indoor spaces.

Astra-Global: The Intelligent Brain for Global Localization

Astra-Global serves as the intelligent core of the Astra architecture, responsible for critical low-frequency tasks: self-localization and target localization. It functions as a Multimodal Large Language Model (MLLM), adept at processing both visual and linguistic inputs to achieve precise global positioning within a map. Its strength lies in utilizing a hybrid topological-semantic graph as contextual input, allowing the model to accurately locate positions based on query images or text prompts.

The construction of this robust localization system begins with offline mapping. The research team developed an offline method to build a hybrid topological-semantic graph G=(V,E,L):

• V (Nodes): Keyframes, obtained by temporal downsampling of input video and SfM-estimated 6-Degrees-of-Freedom (DoF) camera poses, act as nodes encoding camera poses and landmark references.
• E (Edges): Undirected edges establish connectivity based on relative node poses, crucial for global path planning.
• L (Landmarks): Semantic landmark information is extracted by Astra-Global from visual data at each node, enriching the map’s semantic understanding. These landmarks store semantic attributes and are connected to multiple nodes via co-visibility relationships.

In practical localization, Astra-Global’s self-localization and target localization capabilities leverage a coarse-to-fine two-stage process for visual-language localization. The coarse stage analyzes input images and localization prompts, detects landmarks, establishes correspondence with a pre-built landmark map, and filters candidates based on visual consistency. The fine stage then uses the query image and coarse output to sample reference map nodes from the offline map, comparing their visual and positional information to directly output the predicted pose.

For language-based target localization, the model interprets natural language instructions, identifies relevant landmarks using their functional descriptions within the map, and then leverages landmark-to-node association mechanisms to locate relevant nodes, retrieving target images and 6-DoF poses.

To empower Astra-Global with robust localization abilities, the team employed a meticulous training methodology. Using Qwen2.5-VL as the backbone, they combined Supervised Fine-Tuning (SFT) with Group Relative Policy Optimization (GRPO). SFT involved diverse datasets for various tasks, including coarse and fine localization, co-visibility detection, and motion trend estimation. In the GRPO phase, a rule-based reward function (including format, landmark extraction, map matching, and extra landmark rewards) was used to train for visual-language localization. Experiments showed GRPO significantly improved Astra-Global’s zero-shot generalization, achieving 99.9% localization accuracy in unseen home environments, surpassing SFT-only methods.

Astra-Local: The Intelligent Assistant for Local Planning

Astra-Local acts as the intelligent assistant for Astra’s high-frequency tasks, a multi-task network capable of efficiently generating local paths and accurately estimating odometry from sensor data. Its architecture comprises three core components: a 4D spatio-temporal encoder, a planning head, and an odometry head.

The 4D spatio-temporal encoder replaces traditional mobile stack perception and prediction modules. It begins with a 3D spatial encoder that processes N omnidirectional images through a Vision Transformer (ViT) and Lift-Splat-Shoot to convert 2D image features into 3D voxel features. This 3D encoder is trained using self-supervised learning via 3D volumetric differentiable neural rendering. The 4D spatio-temporal encoder then builds upon the 3D encoder, taking past voxel features and future timestamps as input to predict future voxel features through ResNet and DiT modules, providing current and future environmental representations for planning and odometry.

The planning head, based on pre-trained 4D features, robot speed, and task information, generates executable trajectories using Transformer-based flow matching. To prevent collisions, the planning head incorporates a masked ESDF loss (Euclidean Signed Distance Field). This loss calculates the ESDF of a 3D occupancy map and applies a 2D ground truth trajectory mask, significantly reducing collision rates. Experiments demonstrate its superior performance in collision rate and overall score on out-of-distribution (OOD) datasets compared to other methods.

The odometry head predicts the robot’s relative pose using current and past 4D features and additional sensor data (e.g., IMU, wheel data). It trains a Transformer model to fuse information from different sensors. Each sensor modality is processed by a specific tokenizer, combined with modality embeddings and temporal positional embeddi…

Konten dipersingkat otomatis.

🔗 Sumber: syncedreview.com

📌 MAROKO133 Breaking ai: US to convert retired coal mine into 350-megawatt nuclear

The Tennessee Valley Authority (TVA) recently announced that it has issued a Letter of Intent (LOI) to the nuclear fusion company Type One Energy. 

The agreement concerns the potential development of a commercial fusion power plant at the site of the decommissioned Bull Run Fossil Plant near Knoxville, Tennessee.

The proposal focuses on Type One Energy’s Infinity Two power plant design, a 350 megawatt-electric (350 MWe) facility intended to provide baseload power to the grid. The companies are aiming to have the pilot fusion plant operational by the mid-2030s.

Utilizing stellarator fusion technology

The Infinity Two plant is designed to utilize stellarator fusion technology. Unlike the more widely known tokamak design, which uses a toroidal, or doughnut-shaped, chamber, a stellarator employs a complex, twisted magnetic field in a non-symmetrical configuration to contain the superheated plasma necessary for fusion. 

This approach gets around certain plasma confinement challenges faced by tokamaks, which arise from variations in magnetic coil density around the toroidal ring.

In its announcement, TVA noted that the stellarator is currently the only fusion technology to have demonstrated stable, steady-state operation with high efficiency. 

These characteristics are crucial for a power source intended to provide reliable, on-demand electricity at a competitive price. 

“Type One Energy is developing Infinity Two using today’s existing materials and fundamental fusion technologies to support near-term deployment of the technology,” highlighted the fusion company.

Selecting Bull Run site is notable

As a former fossil fuel plant, the location already possesses critical infrastructure, including high-capacity grid connections and access to cooling water from the Clinch River. 

“I am excited about the possibility of the first US commercial stellarator fusion power plant being built in the Tennessee Valley,” remarked Don Moul, TVA President and CEO.

Repurposing such sites is an increasingly common strategy in the energy sector for transitioning to new power sources. 

For a large utility like TVA, the exploration of fusion is part of a long-term strategy to identify reliable, on-demand, and non-emitting power sources that can operate alongside intermittent renewables like solar and wind.

Expansion of the previous contract

Earlier this year, TVA and Type One Energy signed a cooperative agreement to develop initial plans jointly. This was followed in July by the first set of commercial contracts under an initiative called “Project Infinity.” 

As part of these contracts, TVA’s Power Service Shops in Muscle Shoals, Alabama, are assisting in the development of tailored welding and fabrication techniques for a smaller stellarator testbed known as Infinity One. The manufacturing and construction methods developed for this prototype will then be applied to the building of the full-scale Infinity Two plant.

Beyond construction and power generation, the partnership also addresses the potential future use of the prototype facilities as a center for workforce training, preparing the specialized technicians and operators needed for the commercial fusion industry.

“This LOI with TVA, and the role it establishes for Type One Energy as the fusion technology provider for their Infinity Two project, is therefore fully aligned with our goal of pursuing the lowest-risk approach to commercializing fusion energy,” concluded Christofer Mowry, Type One Energy’s CEO.

🔗 Sumber: interestingengineering.com