Posts

Showing posts with the label autonomous vehicles

Time-dependent optical delay lines/ dynamic modulation integrated onto silicon-based photonic chips avoids electronic memory problems

https://www.eurekalert.org/news-releases/1137512 "temporally plastic, foundry-compatible, real-time photonic neural network incorporating time-dependent optical operators: silicon nitride/ silicon-on-insulator with micro-transfer-printed semiconductor optical amplifiers... stable reconfigurable optical kernel, dynamic pulse-level electro-optic modulation, recursive optical delay line update weights... slashes: compute latency 100X, energy consumption 16X, order-of-magnitude reduction in trajectory prediction error for autonomous navigation"
Post a Comment
Read more

Analog in-memory computing systems remaining mathematically precise/ robust against hardware defects

https://techxplore.com/news/2026-07-method-fault-tolerant-analog-memory.html "analog microchip hardware has structural variations/ random background noise, overcome introducing fault-tolerant error-mitigation dynamically detecting/ isolating hardware defects during operation... scales non-Von Neumann AI processors, wraps targeted digital correction loops around analog computing arrays preserving algorithmic precision without complex/ expensive cleanroom fabrication... stable, power-efficient analog AI processors inside power-constrained edge devices/ autonomous vehicles" Related: Innovative design unlocks resilient semiconductor materials https://techxplore.com/news/2026-08-resilient-semiconductor-materials.html
Post a Comment
Read more

Biomimetic titanium dioxide photomemristor vision system adapts to extreme, fluctuating, and mixed lighting conditions in seconds

https://www.eurekalert.org/news-releases/1130994 "computer vision in high-contrast/ rapidly shifting lighting, such as dark nighttime street illuminated by oncoming high-beam headlights, unsafe for autonomous transport, overcome using titanium dioxide/ flexible conductive polymer blend photomemristor... absorbs/ releases surrounding moisture based on environmental light exposure, regulating optical sensitivity within seconds... 4X4 sensor grid/ machine learning identified complex letter patterns in high-contrast environments with over 95% accuracy"
Post a Comment
Read more

Matte/ shiny differentiation by machine vision enhances robotic surgery, industrial inspection, autonomous vehicle glare navigation

https://www.eurekalert.org/news-releases/1128523 "3D sensors can't accurately map mix of matte surfaces (like fabrics, walls)/ highly reflective, specular surfaces (like glass, mirrors, wet surgical tissue)... overcome using deflectometry: algorithm/ laser scanner instantly maps dull, matte object 3D shapes/ repurposes as collective virtual screen, then tracks how light reflects off shiny objects onto this surrounding virtual screen, computing precise shapes of reflective surfaces... moving, mixed-reflectance scene, real time high-resolution 3D video... highly scalable"
Post a Comment
Read more

Ultralow-noise microwave/ millimeter-wave signal generation using a stabilized microcomb-based photonic chip

https://www.eurekalert.org/news-releases/1127576 "as frequency increases signals become more unstable/ exhibit phase noise, optical signals more stable but frequencies too high for electronic circuits... overcome: synchronizing chip-scale microcomb to ultra-stable optical reference making signal 100 million times cleaner/ also using Perfect Soliton Crystal state of light keeping timing jitter at record 3 fs... 6G with more data on same spectrum, ultra-fast, low-latency... compact autonomous vehicle radars distinguishing with higher certainty... mobile atomic clocks, sensors"
Post a Comment
Read more

Packing antennas tightly on a silicon photonic chip without the catastrophic interference that typically ruins solid-state LiDAR signals

https://phys.org/news/2026-05-photonics-advance-enable-compact-high.html "LiDAR sensor needs wide field of view, placing many tiny antennas close together, achieve high-resolution 3D map, but crosstalk jumbling light beam/ spacing farther apart creates ghost beam copies confusing direction... overcome varying corrugations for unique propagation coefficients on each so they ignore each other... coupling from nearly 100% to 1%... LiDAR solid-state chip integrated into vehicle skin/ tiny robot (was spinning bucket on car roof)... scalable CMOS manufacturing, level 5 vehicles" Related: New driving model predicts split-second crash avoidance with humanlike accuracy https://phys.org/news/2026-06-humanlike-accuracy.html
Post a Comment
Read more

Highly optimized/ structurally specialized, instead of bigger, models for faster more energy-efficient AI and autonomous robots/ vehicles

https://phys.org/news/2026-05-tiny-insect-brain-discovery-blueprint.html "AI processes information linearly requiring significant computational power to weigh every variable, overcome mimicking honeybees' complex/ parallel processed decisions using fewer than 1 million neurons identifying potential targets while enabling ms-response brake, ignoring irrelevant data instantly/ focusing only on high-probability successes... drastic energy reduction, reduced data dependency, real-time autonomy... edge AI in small: untethered drones, autonomous vehicles, exploration robots"
Post a Comment
Read more

Securing vehicular ad hoc networks with blockchain-based federated learning to form infrastructure layer of autonomous mobility

https://www.eurekalert.org/news-releases/1124502 "vehicles share speed, location, road condition data to prevent accidents/ optimize flow, but privacy/ security risks, overcome: vehicles train local AI model on own data on own edge computer, only sharing model updates (mathematical weights, not private raw data), blockchain records/ verifies model updates ensuring aren't being hacked, assigns trust score to each vehicle filtering out malicious ones... immune to single point of failure attacking traffic grid... reduces smart city infrastructure's massive bandwidth costs"
Post a Comment
Read more

Tackling optical neural networks' interference enables efficient, low power, speed of light diffractive edge-AI multi-object recognition

https://www.eurekalert.org/news-releases/1120604 "computers using structured light diffraction through layered materials have interference/ cross-talk among multiple objects, better design: reduces cross-talk between different objects, improves simultaneously present multiple target robustness, optimizes how light propagates/ processed across diffractive layers engineered to encode features of different objects... processing happens via light propagation (no electronics required during inference)... autonomous systems, surveillance, biomedical imaging" Related: The quantum key to seeing through chaos https://phys.org/news/2026-05-quantum-key-chaos.html
Post a Comment
Read more

Grounding AI's learning rate in physical, slow-moving oxygen gradient dynamics for faster, more stable/ energy efficient learning

https://techxplore.com/news/2026-04-memristor-built-oxygen-gradient-stability.html "memristor abrupt/ unstable state changes difficult for complex AI tasks like reinforcement learning... overcome with 2nd-order memristor using stable, built-in oxygen gradient using molecularly coordinated layer... creates dynamic barrier evolving very slowly (taking >100 seconds to shift) allowing information to process/ integrate longer... reduced required training iterations 68.75% (static environments)/ 35.65% (dynamic environments), -98.1% conductance modulation for highly granular synaptic weights' control"
Post a Comment
Read more

Spiking neural networks utilizing event-based processing enables energy efficiency in AI-driven autonomous devices

https://techxplore.com/news/2026-03-brain-ai-hardware-autonomous-devices.html "GPUs when running AI always on, even without relevant events occurring, overcome replacing with brain inspired chip minimizing energy drain, firing (spiking) when changed environment detected... learns real-time on chip itself, without needing Wi-Fi or data center... processes navigation using 100 to 1,000X less energy... sub-250g drone flight time hours (was minutes), tiny search-rescue/ industrial inspection robots can run for weeks on single small battery, smart glasses, health-monitoring sensors" Related: Brain-inspired AI is capable of flexible planning and problem-solving https://www.eurekalert.org/news-releases/1137651
Post a Comment
Read more

Reconfigurable all-optical image processing uses rotation/ lateral displacement to dynamically manipulate light in optical devices

https://www.eurekalert.org/news-releases/1111763 "double-layer liquid crystal processes information using light propagation's laws, not digital conversion, switching between twisted/ non-twisted states (synthetic phase modulation)... high-speed, low-power, integrates 8 image processing functions into single architecture: single /dual-channel bright-field imaging, vortex filtering, 1D edge enhancement, vertex recognition, resolution-tunable edge extraction based on photonic spin Hall effect... front-end processor for autonomous driving, biomedical imaging"
Post a Comment
Read more

GNSS-only positioning for autonomous systems in urban canyons sub-meter accurate in environments where GPS/ GNSS methods fail

https://techxplore.com/news/2026-01-gnss-method-stable-positioning-autonomous.html "bypasses integer ambiguity resolution which determines number of radio wavelengths between satellite/ receiver, which causes interference in cities due to signal reflections... overcome using Rao-Blackwellized particle filter probabilistically estimating position: phase likelihood based on carrier-phase measurements'  fractional parts, velocity integration using Kalman filter to process raw Doppler measurements, outlier resilience removing non-Gaussian errors... 30% better than traditional... autonomous vehicles/ robotics" Related: A new route to faster PPP-AR https://www.eurekalert.org/news-releases/1123517 Coloring the map to reduce visual drift in GNSS-denied navigation https://www.eurekalert.org/news-releases/1126513 New positioning framework cuts satellite navigation convergence time from minutes to seconds https://www.eurekalert.org/news-releases/1133059 New hybrid positioning system...
Post a Comment
Read more

Autonomous vehicle algorithms reduce overall traffic risk 26.3%, pedestrian/ cyclists potential harm by 51.7%,, damage to AVs by 8.3%

https://www.eurekalert.org/news-releases/1086924 "considers interactions among multiple road users, social sensitivity... integrates: (1) Individual Risk Assessment for each road user, assessing speed, distances from one another, behavioral predictability, (2) Socially Weighted Risk Mapping, (3) Behavioral Belief Encoding predicting action's affect on overall traffic situation... evaluated using 2K benchmark traffic scenarios... better operational efficiency, driving tasks 13.9% faster on average... flexible/ adaptable to different regulations/ social norms to drive like locals" Related: Vehicles professionally driven from remote stations, using real-time video feeds/ simulator-style https://www.dagens.com/technology/the-technology-that-could-make-owning-a-car-obsolete
Post a Comment
Read more

Inexpensive PINMAP LiDAR detects glass walls as well as traditional LiDAR does, while lowering autonomous driving cost by 90%

https://www.eurekalert.org/news-releases/1085772 "service robots' LiDAR determines distance/ structure projecting light/ measuring reflection time, expensive/ complex, overcome using inexpensive Probabilistic Incremental Navigation-based mapping LiDAR accumulating rare points to probabilistically calculate likelihood of glass wall presence over time, based on map charting/ navigation software, not hardware, improve inexpensive LiDAR sensors' detection performance... 96.77% accuracy... glass/ acrylic inside hospitals, airports, shopping malls, warehouses" Related: Team achieves 3D recognition of transparent objects in less than two seconds https://techxplore.com/news/2025-06-team-3d-recognition-transparent-seconds.html Infrared image sensor for autonomous driving sees clearly even in the fog https://www.eurekalert.org/news-releases/1125028
Post a Comment
Read more

AI-powered single monocular camera enables autonomous vehicles to react fast, see in 3D, and operate more safely/ cheaply

https://techxplore.com/news/2025-03-ai-camera-tech-cars.html "MonoFG: state-of-the-art accuracy, about $300/ unit (LiDAR costs up to $75K)... accurately distinguishea between foreground/ background, detecting pedestrians, cyclists, other vehicles... virtually real-time 18 fps decision-making, accuracy comparable with LiDAR, ranked No.1 for cyclist detection... this human-first AI reduces accident risks and increases public trust in self-driving technology. Real-world testing shows that MonoFG enhances road safety by enabling smarter, real-time decision-making" Related: Japanese Automaker Nissan Is Developing Vehicles With Self-Driving Technology https://www.usnews.com/news/business/articles/2025-09-21/japanese-automaker-nissan-is-developing-self-driving-technology Can shape priors make road perception more reliable for autonomous driving? https://www.eurekalert.org/news-releases/1110920 Hyundai Motor and Kia Unveil ‘Vision Pulse,’ Driver Safety Technology that Detects Beyond ...
Post a Comment
Read more