MAROKO133 Eksklusif ai: Miracle polymer promises to make room-temperature quantum devices

📌 MAROKO133 Update ai: Miracle polymer promises to make room-temperature quantum d

Imagine a world where quantum devices don’t need to hide inside bulky refrigerators colder than outer space. For decades, this has been the biggest roadblock as quantum states vanish unless they are locked in crystals or machines frozen near absolute zero. 

This makes quantum applications impractical for real-world use. Interestingly, a team of researchers from Georgia Institute of Technology and the University of Alabama has found a solution to this problem.

They have developed a new type of polymer, a plastic-like material, that can hold and manipulate quantum states in solid form at room temperature. This achievement could change the way we think about building future quantum devices, bringing them out of extreme lab environments and into everyday use.

Achieving the quantum impossible

For developing a room-temperature quantum material, instead of using rigid crystals like diamond or silicon carbide, the researchers turned to chemistry for answers. They designed a conjugated polymer, a long molecular chain made up of alternating building blocks that conduct electrons. 

One of the blocks was a donor unit based on an organic compound called dithienosilole, and the other was an acceptor unit called thiadiazoloquinoxaline. Together, these units created the right conditions for unpaired electron spins to move along the backbone of the polymer without quickly losing their quantum information.

They placed a silicon atom at the heart of the donor unit. This caused the polymer chain to twist slightly, which prevented the chains from stacking too tightly. Normally, close stacking makes spins interact too strongly, wiping out their delicate quantum states. However, here, the twist reduced those harmful interactions while still allowing electrons to communicate along the chain. 

Next, to make the polymer processable, the researchers attached long hydrocarbon side chains. These side chains kept the molecules from clumping together, ensured the material dissolved easily, and helped maintain electronic coherence across the chain. The researchers then used a mix of theoretical modeling and experiments to confirm that their design worked. 

Simulations showed that as the polymer chain grew longer, the spin density spread out across it. Eventually, the system settled into a high-spin ground state, a low-energy arrangement with two unpaired electrons aligned in the same direction. This type of state is similar to those used in solid-state qubits.

Experimental verification of the material

To validate the results from their simulations in the lab, the researchers first ran magnetometry tests. These showed that the material’s spins behaved as if there were two unpaired electrons aligned in the same direction, a state known as a triplet ground state.

They then used a technique called electron paramagnetic resonance (EPR) spectroscopy. In simple terms, EPR works a bit like MRI but for electrons. It uses microwaves and a magnetic field to catch the tiny magnetic signals of unpaired electrons. 

The results showed narrow and symmetric signals, which is a good sign because it means the spins are behaving in an orderly way. The researchers also measured the g-factor, a number that tells how strongly an electron responds to a magnetic field. 

For a perfectly free electron, the g-factor is about 2.0. The polymer’s g-factor was very close to this value, which means the electrons were not heavily disturbed by their surroundings. This low level of disturbance, called low spin–orbit coupling, helps the quantum states stay stable for longer.

However, the real breakthrough came when they measured how long the spins could remain stable. At room temperature, the polymer’s spin-lattice relaxation time (T1) was about 44 microseconds, and its phase memory time (Tm) was 0.3 microseconds. These values are already better than many other molecular systems. 

When cooled to 5.5 kelvin, T1 jumped to 44 milliseconds and Tm stretched to more than 1.5 microseconds. Most importantly, these results were achieved without embedding the material in frozen solvents or isolating it in special matrices, conditions that usually make molecular systems impractical for real-world devices.

The team also showed that the polymer could undergo Rabi oscillations, a sign of controlled quantum operations. By applying microwave pulses, they could predictably flip the spin states, essentially performing the basic actions needed for quantum computing. 

Finally, they demonstrated that this polymer isn’t just a lab gimmick. It can be made into thin films, works as a p-type semiconductor in transistors, and operates stably under repeated use. This means it can be integrated into electronic devices, combining both charge and spin functions.

An important step for making quantum applications practical 

This discovery is significant because it shows that quantum materials don’t have to be fragile crystals trapped in cryogenic chambers. Instead, they can be flexible, tunable, and processable polymers that still support quantum coherence. 

“This work demonstrates a fundamentally new approach toward practically applicable organic, high-spin qubits that enable coherent control in the solid-state,” the study authors note.

Such materials could open the door to practical quantum sensors that work in everyday conditions, thin-film devices that combine classical electronics with quantum capabilities, and scalable platforms for exploring quantum computing.

However, this innovation doesn’t solve all the challenges associated with quantum computing. For instance, the phase memory time (duration up to which quantum states are in sync) at room temperature is still relatively short compared to what’s needed for large-scale quantum computing. 

The researchers now plan to further optimize the structure, test new donor-acceptor combinations, and explore device architectures where electronic and spin functions can work together.

The study is published in the journal Advanced Materials.

🔗 Sumber: interestingengineering.com

📌 MAROKO133 Eksklusif ai: ByteDance Introduces Astra: A Dual-Model Architecture fo

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