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

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

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 Apache attack helicopter shoots new 30mm rounds to blo

The US Army has live-tested an AH-64 Apache attack helicopter firing the 30×113mm XM1225 Aviation Proximity Explosive (APEX) round to counter drones.

During the trial, the helicopter engaged air-to-air targets at varying ranges. The test highlighted the munition’s accuracy, versatility, and effectiveness against aerial threats, advancing rotary-wing counter-drone capability.

The Apache, primarily designed for anti-armor missions using weapons such as the Joint Air-to-Ground Missile and Hellfire variants, also employs its M230 Chain Gun for engagements against light armor and personnel.

Recently, US Army soldiers evaluated a heavy-lift cargo drone at Fort Stewart, Georgia, as the service considers expanding autonomous resupply capabilities to frontline units.

Apache ammo upgrade

The AH-64 Apache attack helicopter completed a successful live-fire test at Yuma Proving Ground in Arizona.

In December 2025, the Apache carried out its first air-to-air engagement using 30mm proximity-fused ammunition against unmanned aircraft systems (UAS) at different distances. The test showed the round’s accuracy, flexibility, and effectiveness against aerial targets.

The XM1225 APEX round was developed by Product Manager Medium Caliber Ammunition at Picatinny Arsenal, New Jersey. It is designed to defeat modern threats such as drones, exposed personnel, and small boats.

Importantly, it can be used without any changes to the Apache’s M230 Area Weapon System or its fire control system, allowing easy integration into existing aircraft.

According to the US Army, the ammunition has undergone extensive safety testing to ensure reliable performance. During the trial, the main goal was to measure the XM1225’s accuracy and compare it with the older M789 High Explosive Dual Purpose round under the same conditions.

A secondary goal was to gather data on firing mixed loads of XM1225 and M789 rounds against both ground and drone targets.

Counter-drone firepower

Initial test results showed the XM1225 met all accuracy requirements and proved highly effective against both ground targets and unmanned aircraft systems.

The US Army claims its proximity fuze allows the round to detonate near a target rather than on direct impact, creating a wider lethal radius. This increases its ability to defeat airborne threats and dispersed ground targets, improving overall battlefield effectiveness. The capability enhances the Apache’s performance in both air-to-ground and air-to-air engagements, strengthening its role in modern combat operations.

Test officials noted that the proximity fuze can significantly increase the vulnerability of lightly protected ground and aerial targets, provided they can be properly detected, identified, and tracked. The round’s effectiveness depends on accurate targeting, but it offers expanded engagement options once those conditions are met.

The XM1225 APEX round also maintains similar ballistic characteristics to the currently fielded M789 High Explosive Dual Purpose round. This similarity allows it to be integrated into existing platforms without major adjustments. As a result, the new ammunition adds greater lethality to the Apache without requiring significant additional training for pilots or maintenance personnel.

“Designed to counter emerging threats with unmatched precision and lethality, the XM1225 adds a new capability to the arsenal of 30mm proximity ammunition, giving the Apache Attack Helicopter another lethal option to hunt and defeat modernized threats,” said the US Army in a statement.

🔗 Sumber: interestingengineering.com