arXiv subjects

Ashutosh Mishra

Publications and source records attributed to Ashutosh Mishra.

At least 19 recordsLinked to original sources

arXiv2026

Analytical Series Expansion for Efficient Gradient Evaluation in Multi-Qubit Optimal Control

The open-loop optimization of quantum dynamics using gradient-based quantum optimal control methods involves calculating the time-ordered propagator and its gradient. In this Letter, we present a unifying framework for gradient-based quantum optimal control with respect to any general pulse parameterization by deriving the formal solution from first principles. For the case of unitary propagators, we derive a series expansion involving time-independent commutators and time-dependent coefficients, significantly reducing the number of matrix exponentials needed to compute the gradient. The expansion highlights the connection between derivatives of the propagator and operator evolution in the Heisenberg picture. The method is particularly suited for simulating optimal control tasks in quantum systems with local interactions, which is a common situation in large multi-qubit platforms. We compare the computational cost required for the series with the Gradient Optimization of Analytic conTrols (GOAT) method, and, focusing on the problem of preparation of a GHZ state, demonstrate more than an order of magnitude speedup for a qubit ladder and a chain geometry.

quant-ph↗

arXiv2026

Relative Positions Generalize, Absolute Positions Memorize: An Implicit-Bias Account of Length Generalization in Attention

Transformers with relative positional encodings often extrapolate to sequences longer than those seen during training, whereas transformers with learned absolute encodings typically do not. This is a robust empirical regularity, and the explanations offered for it so far are chiefly about expressivity, that is, about whether a length-generalizing solution exists. We give an optimization explanation. On a minimal fixed-offset retrieval task that isolates positional selection, the gap is governed by the implicit bias of the trained attention head: among the many solutions that fit short sequences, which one gradient descent actually selects. We prove that rotary encodings make the attention logit a function of relative offset alone, an exact equivariance, so whatever selection rule is learned at training lengths is reproduced verbatim at every longer length. Learned absolute encodings instead leave out-of-range positions unconstrained, and the trained head pins to a fixed absolute position inside the training range. We characterize the learned rotary rule as a low-rank ``carrier'' kernel aligned with the target offset, and we derive the resulting graceful accuracy decay as an attention-dilution law; both predictions are confirmed across seeds and offsets. A linear-attention control shows the mechanism is specific to softmax: without normalization, training selects a min-norm interpolant that does not extrapolate. The phenomenon, the equivariance, and the carrier all transfer to a multi-layer, multi-head transformer trained on a full-sequence length-generalization task. The account connects the implicit bias of attention, implicit bias for extrapolation in recurrent models, and the learning side of the RASP-L conjecture.

cs.LG↗

arXiv2026

Same Loss, Same Noise, Opposite Schedules: Noise Structure and Optimizer Normalization Jointly Determine Whether Learning-Rate Cooldown Helps

The cooldown phase of a warmup-stable-decay (WSD) learning-rate schedule, now a default in large-model pretraining, lowers the final training loss in some settings and does nothing in others. We give a provable account of which case obtains, and it turns on two properties together: the structure of the gradient noise and whether the optimizer normalizes its update. On a strongly convex objective with multiplicative (gradient-proportional) noise, stochastic gradient descent contracts geometrically at a constant learning rate, so cooldown has nothing to improve. Under the same objective and noise, sign-based and normalized methods, the standard surrogates for adaptive optimizers, settle on a noise floor of order $\eta^2$ and reach the minimizer only as the learning rate is driven to zero; any additive noise then reinstates a floor for every method. The mechanism is elementary: an SGD step shrinks in proportion to the gradient and so anneals itself, whereas a normalized step keeps unit scale and cannot. We solve the signSGD stationary law on the quadratic exactly and obtain the floor constant in closed form, prove a local form of the dissociation under $(L_0,L_1)$-smoothness, extend the floor to normalized SGD in dimension d>1 by a scale-invariance argument, and establish robustness to momentum and heavy-tailed noise. Simulation confirms every prediction, and we demonstrate the resulting noise-regime diagnostic on a real classification task with directly measured gradient noise. The mechanism explains whether cooldown helps; the interior cooldown fraction used at scale lies outside stationary landscape-and-noise geometry.

cs.LG↗

arXiv2026

Distributed Multi Robot Lunar Cargo Transportation via Phase Decomposed Reinforcement Learning

Modular reconfigurable robotic systems provide a scalable solution for cooperative surface operations in future lunar missions. However, cooperative cargo transportation remains challenging due to morphology-dependent topology changes, strong payload-induced coupling, long-horizon decision making, and safety constraints. This paper proposes a phase-decomposed reinforcement learning framework for cooperative cargo transport with distributed robotic units. The task is decomposed into lifting, transportation, and placement, each optimized with a dedicated joint-state policy capturing inter-agent coupling. Centralized training promotes stable convergence, while deployment uses onboard proprioception for control and OptiTrack motion capture for ground-truth evaluation and post-processed metrics. A deterministic phase controller expressed in Markov state representation regulates transitions between stages, and a failure-sensitive synchronization mechanism ensures coordinated progression and safety-aware halting during real-world execution. The framework is evaluated in simulation and through controlled field experiments at a JAXA space exploration test facility. Results demonstrate reliable cooperative transport across all stages in both simulation and hardware experiments.

cs.RO↗

arXiv2025

MoonBot: Modular and On-Demand Reconfigurable Robot Toward Moon Base Construction

The allure of lunar surface exploration and development has recently captured widespread global attention. Robots have proved to be indispensable for exploring uncharted terrains, uncovering and leveraging local resources, and facilitating the construction of future human habitats. In this article, we introduce the modular and on-demand reconfigurable robot (MoonBot), a modular and reconfigurable robotic system engineered to maximize functionality while operating within the stringent mass constraints of lunar payloads and adapting to varying environmental conditions and task requirements. This article details the design and development of MoonBot and presents a preliminary field demonstration that validates the proof of concept through the execution of milestone tasks simulating the establishment of lunar infrastructure. These tasks include essential civil engineering operations, infrastructural component transportation and deployment, and assistive operations with inflatable modules. Furthermore, we systematically summarize the lessons learned during testing, focusing on the connector design and providing valuable insights for the advancement of modular robotic systems in future lunar missions.

cs.RO↗

arXiv2025

Solving the Cosmological Vlasov-Poisson Equations with Physics-Informed Kolmogorov-Arnold Networks

Cold dark matter (CDM) evolves as a collisionless fluid under the Vlasov-Poisson equations, but N-body simulations approximate this evolution by discretising the distribution function into particles, introducing discreteness effects at small scales. We present a physics-informed neural network approach that evolves CDM fields without any use of N-body data or methods, using a Kolmogorov-Arnold network (KAN) to model the continuous displacement field for one-dimensional halo collapse. Physical constraints derived from the Vlasov-Poisson equations are embedded directly into the loss function, enabling accurate evolution beyond the first shell crossing. The trained model achieves sub-percent errors on the residuals even after seven shell crossings and matches N-body results while providing a resolution-free representation of the displacement field. In addition, displacement errors do not grow over time, a very interesting feature with respect to N-body methods. It can also reconstruct initial conditions through backward evolution when sufficient final-state information is available. Although current runtimes exceed those of N-body methods, this framework offers a new route to high-fidelity CDM evolution without particle discretisation, with prospects for extension to higher dimensions.

astro-ph.CO↗

arXiv2025

Designing for Distributed Heterogeneous Modularity: On Software Architecture and Deployment of MoonBots

This paper presents the software architecture and deployment strategy behind the MoonBot platform: a modular space robotic system composed of heterogeneous components distributed across multiple computers, networks and ultimately celestial bodies. We introduce a principled approach to distributed, heterogeneous modularity, extending modular robotics beyond physical reconfiguration to software, communication and orchestration. We detail the architecture of our system that integrates component-based design, a data-oriented communication model using ROS2 and Zenoh, and a deployment orchestrator capable of managing complex multi-module assemblies. These abstractions enable dynamic reconfiguration, decentralized control, and seamless collaboration between numerous operators and modules. At the heart of this system lies our open-source Motion Stack software, validated by months of field deployment with self-assembling robots, inter-robot cooperation, and remote operation. Our architecture tackles the significant hurdles of modular robotics by significantly reducing integration and maintenance overhead, while remaining scalable and robust. Although tested with space in mind, we propose generalizable patterns for designing robotic systems that must scale across time, hardware, teams and operational environments.

cs.RO↗

arXiv2025

Design and Development of a Modular Bucket Drum Excavator for Lunar ISRU

In-Situ Resource Utilization (ISRU) is one of the key technologies for enabling sustainable access to the Moon. The ability to excavate lunar regolith is the first step in making lunar resources accessible and usable. This work presents the development of a bucket drum for the modular robotic system MoonBot, as part of the Japanese Moonshot program. A 3D-printed prototype made of PLA was manufactured to evaluate its efficiency through a series of sandbox tests. The resulting tool weighs 4.8 kg and has a volume of 14.06 L. It is capable of continuous excavation at a rate of 777.54 kg/h with a normalized energy consumption of 0.022 Wh/kg. In batch operation, the excavation rate is 172.02 kg/h with a normalized energy consumption of 0.86 Wh per kilogram of excavated material. The obtained results demonstrate the successful implementation of the concept. A key advantage of the developed tool is its compatibility with the modular MoonBot robotic platform, which enables flexible and efficient mission planning. Further improvements may include the integration of sensors and an autonomous control system to enhance the excavation process.

cs.RO↗

arXiv2025

Multi-Modal Decentralized Reinforcement Learning for Modular Reconfigurable Lunar Robots

Modular reconfigurable robots suit task-specific space operations, but the combinatorial growth of morphologies hinders unified control. We propose a decentralized reinforcement learning (Dec-RL) scheme where each module learns its own policy: wheel modules use Soft Actor-Critic (SAC) for locomotion and 7-DoF limbs use Proximal Policy Optimization (PPO) for steering and manipulation, enabling zero-shot generalization to unseen configurations. In simulation, the steering policy achieved a mean absolute error of 3.63° between desired and induced angles; the manipulation policy plateaued at 84.6 % success on a target-offset criterion; and the wheel policy cut average motor torque by 95.4 % relative to baseline while maintaining 99.6 % success. Lunar-analogue field tests validated zero-shot integration for autonomous locomotion, steering, and preliminary alignment for reconfiguration. The system transitioned smoothly among synchronous, parallel, and sequential modes for Policy Execution, without idle states or control conflicts, indicating a scalable, reusable, and robust approach for modular lunar robots.

cs.RO↗

arXiv2025

Enhancing Autonomous Manipulator Control with Human-in-loop for Uncertain Assembly Environments

This study presents an advanced approach to enhance robotic manipulation in uncertain and challenging environments, with a focus on autonomous operations augmented by human-in-the-loop (HITL) control for lunar missions. By integrating human decision-making with autonomous robotic functions, the research improves task reliability and efficiency for space applications. The key task addressed is the autonomous deployment of flexible solar panels using an extendable ladder-like structure and a robotic manipulator with real-time feedback for precision. The manipulator relays position and force-torque data, enabling dynamic error detection and adaptive control during deployment. To mitigate the effects of sinkage, variable payload, and low-lighting conditions, efficient motion planning strategies are employed, supplemented by human control that allows operators to intervene in ambiguous scenarios. Digital twin simulation enhances system robustness by enabling continuous feedback, iterative task refinement, and seamless integration with the deployment pipeline. The system has been tested to validate its performance in simulated lunar conditions and ensure reliability in extreme lighting, variable terrain, changing payloads, and sensor limitations.

cs.RO↗

arXiv2025

Robust and Modular Multi-Limb Synchronization in Motion Stack for Space Robots with Trajectory Clamping via Hypersphere

Modular robotics holds immense potential for space exploration, where reliability, repairability, and reusability are critical for cost-effective missions. Coordination between heterogeneous units is paramount for precision tasks -- whether in manipulation, legged locomotion, or multi-robot interaction. Such modular systems introduce challenges far exceeding those in monolithic robot architectures. This study presents a robust method for synchronizing the trajectories of multiple heterogeneous actuators, adapting dynamically to system variations with minimal system knowledge. This design makes it inherently robot-agnostic, thus highly suited for modularity. To ensure smooth trajectory adherence, the multidimensional state is constrained within a hypersphere representing the allowable deviation. The distance metric can be adapted hence, depending on the task and system under control, deformation of the constraint region is possible. This approach is compatible with a wide range of robotic platforms and serves as a core interface for Motion-Stack, our new open-source universal framework for limb coordination (available at https://github.com/2lian/Motion-Stack ). The method is validated by synchronizing the end-effectors of six highly heterogeneous robotic limbs, evaluating both trajectory adherence and recovery from significant external disturbances.

cs.RO↗

arXiv2025

Demolition measurement protocol for transmon qubits

The process of measuring a qubit and re-initializing it to the ground state practically lead to long qubit idle times between re-runs of experiments on a superconducting quantum computer. Here, we propose a protocol for a \textit{demolition measurement} of a transmon qubit that integrates qubit readout with the reset process to minimize qubit idle time. We present a three-staged implementation of this protocol, involving a combined qubit readout and resonator reset scheme that unconditionally resets the resonator at the end of the readout; a leakage removal scheme that can be integrated with the measurement stage; and an unconditional qubit reset. We demonstrate that this protocol could be implemented in $1 \, μ$s with greater than $95 \, \%$ reset fidelity and a $99 \, \%$ readout fidelity without any hardware overhead beyond those commonly used. This provides at least a 50x speed up compared to the passive decay of the qubit, thereby significantly increasing the data-acquisition rate.

quant-ph↗

arXiv2024

Combining summary statistics with simulation-based inference for the 21 cm signal from the Epoch of Reionization

The 21 cm signal from the Epoch of Reionization will be observed with the up-coming Square Kilometer Array (SKA). SKA should yield a full tomography of the signal which opens the possibility to explore its non-Gaussian properties. How can we extract the maximum information from the tomography and derive the tightest constraint on the signal? In this work, instead of looking for the most informative summary statistics, we investigate how to combine the information from two sets of summary statistics using simulation-based inference. To this purpose, we train Neural Density Estimators (NDE) to fit the implicit likelihood of our model, the LICORICE code, using the Loreli II database. We train three different NDEs: one to perform Bayesian inference on the power spectrum, one to do it on the linear moments of the Pixel Distribution Function (PDF) and one to work with the combination of the two. We perform $\sim 900$ inferences at different points in our parameter space and use them to assess both the validity of our posteriors with Simulation-based Calibration (SBC) and the typical gain obtained by combining summary statistics. We find that our posteriors are biased by no more than $\sim 20 \%$ of their standard deviation and under-confident by no more than $\sim 15 \%$. Then, we establish that combining summary statistics produces a contraction of the 4-D volume of the posterior (derived from the generalized variance) in 91.5 % of our cases, and in 70 to 80 % of the cases for the marginalized 1-D posteriors. The median volume variation is a contraction of a factor of a few for the 4D posteriors and a contraction of 20 to 30 % in the case of the marginalized 1D posteriors. This shows that our approach is a possible alternative to looking for sufficient statistics in the theoretical sense.

astro-ph.CO↗

arXiv2024

CoroNetGAN: Controlled Pruning of GANs via Hypernetworks

Generative Adversarial Networks (GANs) have proven to exhibit remarkable performance and are widely used across many generative computer vision applications. However, the unprecedented demand for the deployment of GANs on resource-constrained edge devices still poses a challenge due to huge number of parameters involved in the generation process. This has led to focused attention on the area of compressing GANs. Most of the existing works use knowledge distillation with the overhead of teacher dependency. Moreover, there is no ability to control the degree of compression in these methods. Hence, we propose CoroNet-GAN for compressing GAN using the combined strength of differentiable pruning method via hypernetworks. The proposed method provides the advantage of performing controllable compression while training along with reducing training time by a substantial factor. Experiments have been done on various conditional GAN architectures (Pix2Pix and CycleGAN) to signify the effectiveness of our approach on multiple benchmark datasets such as Edges-to-Shoes, Horse-to-Zebra and Summer-to-Winter. The results obtained illustrate that our approach succeeds to outperform the baselines on Zebra-to-Horse and Summer-to-Winter achieving the best FID score of 32.3 and 72.3 respectively, yielding high-fidelity images across all the datasets. Additionally, our approach also outperforms the state-of-the-art methods in achieving better inference time on various smart-phone chipsets and data-types making it a feasible solution for deployment on edge devices.

cs.CV↗

arXiv2023

Chaurah: A Smart Raspberry Pi based Parking System

The widespread usage of cars and other large, heavy vehicles necessitates the development of an effective parking infrastructure. Additionally, algorithms for detection and recognition of number plates are often used to identify automobiles all around the world where standardized plate sizes and fonts are enforced, making recognition an effortless task. As a result, both kinds of data can be combined to develop an intelligent parking system focuses on the technology of Automatic Number Plate Recognition (ANPR). Retrieving characters from an inputted number plate image is the sole purpose of ANPR which is a costly procedure. In this article, we propose Chaurah, a minimal cost ANPR system that relies on a Raspberry Pi 3 that was specifically created for parking facilities. The system employs a dual-stage methodology, with the first stage being an ANPR system which makes use of two convolutional neural networks (CNNs). The primary locates and recognises license plates from a vehicle image, while the secondary performs Optical Character Recognition (OCR) to identify individualized numbers from the number plate. An application built with Flutter and Firebase for database administration and license plate record comparison makes up the second component of the overall solution. The application also acts as an user-interface for the billing mechanism based on parking time duration resulting in an all-encompassing software deployment of the study.

cs.CV↗

arXiv2023

Size-dependent Failure Behavior of Lithium-Iron Phosphate Battery under Mechanical Abuse

Under mechanical abuse conditions, the failure of lithium-ion batteries occurs in various stages characterized by different force, temperature and voltage response which require it's in-situ measurements for analysis. Firstly, four sizes of commercially available lithium-iron phosphate batteries (LFPB) viz. 18650, 22650, 26650, and 32650 are subjected to quasi static lateral, longitudinal compression, and nail penetration tests. The failure, characterized by the voltage drop and temperature rise, at the onset of the first internal short-circuit (ISC), is identified by Aurdino-based voltage sensor module and temperature measurement module, respectively. The battery failure load and peak temperature at the onset of ISC are found to rely on the battery size strongly. The failure is observed to be delayed for small-sized 18650 batteries during lateral compression, unlike longitudinal compression and nail penetration test. At the onset of the short circuit, the temperature rise above the ambient value is different for different LFPBs. It is found to be maximum (64.4 degree C ) for LFPB 32650 under longitudinal compression and minimum (29.5 degree C) under lateral compression tests amongst the considered geometries. Further, LFPB 26650 exhibited a balanced thermal behavior during the considered abused condition. Such data can be sensed timely for effective thermal management and improved safety of lithium-ion batteries.

cond-mat.mtrl-sci↗

arXiv2022

A Survey on Deep Reinforcement Learning-based Approaches for Adaptation and Generalization

Deep Reinforcement Learning (DRL) aims to create intelligent agents that can learn to solve complex problems efficiently in a real-world environment. Typically, two learning goals: adaptation and generalization are used for baselining DRL algorithm's performance on different tasks and domains. This paper presents a survey on the recent developments in DRL-based approaches for adaptation and generalization. We begin by formulating these goals in the context of task and domain. Then we review the recent works under those approaches and discuss future research directions through which DRL algorithms' adaptability and generalizability can be enhanced and potentially make them applicable to a broad range of real-world problems.

cs.LG↗

arXiv2019

Using Deep Object Features for Image Descriptions

Inspired by recent advances in leveraging multiple modalities in machine translation, we introduce an encoder-decoder pipeline that uses (1) specific objects within an image and their object labels, (2) a language model for decoding joint embedding of object features and the object labels. Our pipeline merges prior detected objects from the image and their object labels and then learns the sequences of captions describing the particular image. The decoder model learns to extract descriptions for the image from scratch by decoding the joint representation of the object visual features and their object classes conditioned by the encoder component. The idea of the model is to concentrate only on the specific objects of the image and their labels for generating descriptions of the image rather than visual feature of the entire image. The model needs to be calibrated more by adjusting the parameters and settings to result in better accuracy and performance.

cs.CV↗