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

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

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 Hot ai: 100-foot tall US nuclear DOME launched for advanced 20 MW micr

The Idaho National Laboratory (INL) has officially opened the Demonstration of Microreactor Experiments (DOME) nuclear test bed. 

This facility is located at the laboratory’s Materials and Fuels Complex and serves as a site for the testing and demonstration of advanced nuclear reactors developed by private companies. 

The test bed was established to provide a location where experimental reactor designs can be evaluated in a controlled environment.

The physical structure of the DOME uses the repurposed Experimental Breeder Reactor-II dome. This building has a diameter of 80 feet and a height of 100 feet. 

By utilizing an existing containment structure, the project was able to establish a testing environment without constructing an entirely new shell.

The facility is designed to support microreactor experiments that generate up to 20 megawatts of thermal energy. This thermal capacity allows for the testing of various reactor concepts at operational temperatures and power levels.

Facilitating the collection of performance data

The primary objective of the DOME is to facilitate the collection of performance data from new reactor designs. Companies use this data to verify their technical calculations and to prepare for the licensing process required by regulatory agencies.

By providing a site with the necessary infrastructure and safety systems, the facility is intended to decrease the amount of time and the financial resources needed to move a nuclear design from a concept to a functional unit. 

As part of a national laboratory, the DOME offers access to technical staff and specialized equipment that are not typically available in the private sector.

Dr. Rian Bahran, Deputy Assistant Secretary for Nuclear Reactors, stated that the test bed is a component of the Department of Energy’s strategy for nuclear technology development. 

The infrastructure provided by the DOME allows for the validation of new designs. According to Bahran, the facility supports national goals related to energy security by providing a location where energy solutions can be tested and prepared for deployment.

John Wagner, the Director of Idaho National Laboratory, noted that the facility is an investment in infrastructure intended to meet the needs of the nuclear industry. He stated that the DOME enables developers to move from the conceptual phase to the demonstration phase. 

For current industrial demands

The laboratory’s goal is to provide a setting where these transitions occur on a timeline that meets current industrial demands.

Brad Tomer, the Director of NRIC, stated that the facility was created to give the industry a place to convert concepts into demonstrations. The data gathered during these tests is used to confirm the functionality of the technologies.

The first scheduled experiment involving nuclear fuel at the DOME is expected to take place later this year. Radiant’s Kaleidos Demonstration Unit is the first project identified for testing at the site. 

The company plans to begin a year-long testing program during the current spring season. This will be the first instance of a fueled reactor being tested within the repurposed facility.

Microreactors are a category of nuclear reactors that are smaller than the units used in traditional power plants. These designs are often intended to be manufactured in a factory and transported to a location for use. 

The DOME provides the shielding and safety controls required to test these smaller units before they are moved to their final destinations.

🔗 Sumber: interestingengineering.com