Australia to spend $1.1bn on underwater'Ghost Shark' attack drones Australia will spend 1.7 billion Australian dollars ($1.1bn) on a fleet of extra-large underwater "Ghost Shark" attack drones, in a move that officials said would supplement the country's plans to acquire sophisticated nuclear-powered submarines. Australian Minister for Defence Richard Marles said on Wednesday that the Ghost Shark autonomous underwater vehicles will complement Australia's naval surface fleet and submarines to provide "a more capable and more lethal navy". "We have consistently articulated that Australia faces the most complex, in some ways, the most threatening, strategic landscape that we have had since the end of the second world war," Marles said. The government said it signed the $1.1bn, five-year contract with Anduril Australia to build, maintain and develop the uncrewed undersea vehicles in Australia. "This is the highest tech capability in the world," Marles said, adding that the drones would have a "very long range" as well as stealth capabilities.

Accurately predicting travel mode choice is essential for effective transportation planning, yet traditional statistical and machine learning models are constrained by rigid assumptions, limited contextual reasoning, and reduced generalizability. This study explores the potential of Large Language Models (LLMs) as a more flexible and context-aware approach to travel mode choice prediction, enhanced by Retrieval-Augmented Generation (RAG) to ground predictions in empirical data. We develop a modular framework for integrating RAG into LLM-based travel mode choice prediction and evaluate four retrieval strategies: basic RAG, RAG with balanced retrieval, RAG with a cross-encoder for re-ranking, and RAG with balanced retrieval and cross-encoder for re-ranking. These strategies are tested across three LLM architectures (OpenAI GPT-4o, o4-mini, and o3) to examine the interaction between model reasoning capabilities and retrieval methods. Using the 2023 Puget Sound Regional Household Travel Survey data, we conduct a series of experiments to evaluate model performance. The results demonstrate that RAG substantially enhances predictive accuracy across a range of models. Notably, the GPT-4o model combined with balanced retrieval and cross-encoder re-ranking achieves the highest accuracy of 80.8%, exceeding that of conventional statistical and machine learning baselines. Furthermore, LLM-based models exhibit superior generalization abilities relative to these baselines. Findings highlight the critical interplay between LLM reasoning capabilities and retrieval strategies, demonstrating the importance of aligning retrieval strategies with model capabilities to maximize the potential of LLM-based travel behavior modeling.

To account also for spatial dependencies, Cini et al.

Breakthroughs, discoveries, and DIY tips sent every weekday. It wasn't quite the eureka moment a team of scientists in Japan had set out for. Leo Uesaka, a marine biologist at the University of Tokyo, planned to study how seabirds use their legs to take flight from the ocean surface. He secured matchbox-sized cameras to the undersides of 15 streaked shearwaters (Calonectris leucomelas), a Pacific Ocean petrel species, to observe their movements. The tiny, tail-facing cameras successfully recorded information on the birds' legs.