lidar

Claude Opus agents autonomously discovered that Washington State DNR's 2024 aerial LIDAR survey of Puget Sound captured a detailed scan of the USS Nimitz, allowing the team to match their Three.js aircraft carrier model to nearly exact real-world dimensions.

This paper presents a sim-to-real transfer approach for Vision-Language Navigation in continuous environments, using a Cross-Modal Attention architecture fine-tuned with real-world data from a custom Ackermann-steered robot equipped with camera and LiDAR, achieving effective offline navigation without relying on navigation graphs or panoramic views.

Flatten SF is a browser-based tool that computes the flattest routes between any two points in San Francisco using 160,000 street segments, USGS 1 m lidar elevation data, and the Overture/OpenStreetMap street graph, with an interactive slider exploring the distance-vs-climbing trade-off. Source code and analysis are published on GitHub.

This paper introduces a modified Stable Diffusion 3 architecture with multimodal conditioning to refine photogrammetric digital surface models, improving elevation accuracy in French cities by reducing noise and errors.

Fei-Fei Li's AI model for 3D reconstruction faces significant challenges in shipping as a reliable product, highlighting the gap between research and real-world application. The article compares it to LiDAR, discussing trade-offs in accuracy and accessibility for industries like construction and design.

Dyson has launched its new Nurovi line of robot vacuums, featuring three models with lidar navigation, UV light for stain detection, and Twin-Capture System for cleaning on various surfaces.

This paper introduces a fully Bayesian framework for point cloud curve reconstruction using Markov chain Monte Carlo algorithms, which handles noise and missing data while providing uncertainty quantification, with experiments on synthetic and real-world LiDAR data showing accurate results.

Waymo argues that cameras alone are insufficient for fully autonomous driving, emphasizing the need for lidar and radar in contrast to Tesla's camera-only approach.

RapidLiDAR introduces a real-time LiDAR scene completion method using adaptive initialization and multi-scale reconstruction, achieving state-of-the-art performance 2.3 times faster to match automotive LiDAR rates.

Presents a globally trained CNN for forest above-ground biomass estimation using multi-sensor satellite data, with a sparse field calibration workflow to adapt predictions locally. Achieves improved accuracy over uncalibrated global models and ESA CCI products.

The FCC has proposed banning foreign-made drones equipped with LiDAR, thermal imaging, or other 'military-grade' features from U.S. sale, which would remove popular DJI models like the Air 3S and Mini 5 Pro from shelves while leaving existing owners unaffected.

Archaeologists using LiDAR discovered a massive ancient Amazonian civilization with 20,000+ earthworks, rewriting assumptions about pre-Columbian history and challenging earlier beliefs about the region's capacity for complex societies.

MIT Media Lab researchers demonstrate that consumer-grade LiDAR sensors (like those in iPhones) can be used to see around corners, using a technique called motion-induced aperture sampling, enabling non-line-of-sight imaging with off-the-shelf hardware for under $100.

Tesla's robotaxi mileage is flat and vehicle count appears shrinking, contrasting with Waymo's much higher mileage, and investors are growing concerned despite Musk's optimistic framing during an earnings call.

LIDAR-AD proposes a decoder-free latent-interaction world model for autonomous driving that uses redundancy-reduced latent alignment and residual action updates to improve risk-aware state abstraction and long-horizon dynamics prediction, outperforming baseline world models in simulated and real-world scenarios.

The article announces oomwoo, an open-source, build-it-yourself robot vacuum cleaner that uses 2D LiDAR and ROS 2 for autonomous navigation, with no cloud required. It is designed for the maker community and is built in public from the start.

A paper proposing FLISP, a mapless path planning framework for cooperative UGV-UAV tunnel inspection using a single LiDAR-IMU suite, achieving 100% success rate and 7ms latency in a 1.2km tunnel.

OmniPath is a multi-modal agentic framework that combines OpenStreetMap network topology with aerial LiDAR data to audit wheelchair accessibility by analyzing physical barriers like slope and surface discontinuities at high resolution, validated against field surveys.

ShotcreteDepth is a bi-modal dataset of stereo RGB and LiDAR data from construction environments, designed to support research in depth perception under challenging conditions. The dataset includes 11,252 samples with 220 annotated, and is accompanied by a lightweight annotation tool.

KITScenes Multimodal is a high-fidelity European autonomous driving dataset with synchronized sensors, complete 3D HD maps, and four benchmarks for spatial learning and embodied AI research.