Search

The Robot Operating System (ROS) is one of the most popular middleware for developing robot applications, but it is subject to major shortcomings when applied to real-time robotic systems in safety-critical environments. For this reason, ROS 2 was released in 2017 for implementing real-time capabilities in distributed robotic systems while supporting the most prominent aspects of the original ROS. There is still not much work done to provide formal guarantees and correctness of a ROS program. In this paper, we propose a framework to address this challenging problem of guaranteeing the correct behaviour of robotic systems. We propose a formal modelling of a ROS 2 program, and also describe the program using a network of timed automata. We then prove that the sets of executions of a ROS program in the model and in the network of timed automata are the same. Thus to analyze a publisher-subscriber scenario of ROS 2 program, our algorithm first converts the program into the model, and then into the network of timed automata. The applicability and validity of our approach are verified by conducting several experiments on a simplified system and an actual robotic system, and the results and limitations are discussed., Comment: The co-authors have asked to withdraw this paper, since it contains incomplete and incorrect informations

Coverage path planning (CPP) is the task of computing an optimal path within a region to completely scan or survey an area of interest using one or multiple mobile robots. Robots equipped with sensors and cameras can collect vast amounts of data on crop health, soil conditions, and weather patterns. Advanced analytics can then be applied to this data to make informed decisions, improving overall farm management. In this paper, we will demonstrate one approach to find the optimal coverage path of an agricultural field using a single robot, and one using multiple robots. For the single robot, we used a wavefront coverage algorithm that generates a sequence of locations that the robot needs to follow. For the multi-robot approach, the proposed approach consists of two steps: dividing the agricultural field into convex polygonal areas to optimally distribute them among the robots, and generating an optimal coverage path to ensure minimum coverage time for each of the polygonal areas., Comment: The co-authors have asked to withdraw this paper, since it contains incomplete and incorrect informations

Machine learning systems are increasingly being used in critical decision making such as healthcare, finance, and criminal justice. Concerns around their fairness have resulted in several bias mitigation techniques that emphasize the need for high-quality data to ensure fairer decisions. However, the role of earlier stages of machine learning pipelines in mitigating model bias has not been explored well. In this paper, we focus on the task of acquiring additional labeled data points for training the downstream machine learning model to rapidly improve its fairness. Since not all data points in a data pool are equally beneficial to the task of fairness, we generate an ordering in which data points should be acquired. We present DataSift, a data acquisition framework based on the idea of data valuation that relies on partitioning and multi-armed bandits to determine the most valuable data points to acquire. Over several iterations, DataSift selects a partition and randomly samples a batch of data points from the selected partition, evaluates the benefit of acquiring the batch on model fairness, and updates the utility of partitions depending on the benefit. To further improve the effectiveness and efficiency of evaluating batches, we leverage influence functions that estimate the effect of acquiring a batch without retraining the model. We empirically evaluate DataSift on several real-world and synthetic datasets and show that the fairness of a machine learning model can be significantly improved even while acquiring a few data points., Comment: 19 pages, 9 figures

This paper presents a comprehensive analysis of quantization techniques for optimizing Large Language Models (LLMs), specifically focusing on Post-Training Quantization (PTQ) and Quantization-Aware Training (QAT). Through empirical evaluation across models ranging from 10M to 1B parameters, we demonstrate that quantization can achieve up to 68% reduction in model size while maintaining performance within 6% of full-precision baselines when utilizing our proposed scaling factor {\gamma}. Our experiments show that INT8 quantization delivers a 40% reduction in computational cost and power consumption, while INT4 quantization further improves these metrics by 60%. We introduce a novel theoretical framework for mixed-precision quantization, deriving optimal bit allocation strategies based on layer sensitivity and weight variance. Hardware efficiency evaluations on edge devices reveal that our quantization approach enables up to 2.4x throughput improvement for INT8 and 3x for INT4, with 60% power reduction compared to full-precision models., Comment: 9 pages, 2 figures

We investigate inverse backward-in-time problems for a class of second-order degenerate Mean-Field Game (MFG) systems. More precisely, given the final datum $(u(\cdot, T),m(\cdot, T))$ of a solution to the one-dimensional mean-field game system with a degenerate diffusion coefficient, we aim to determine the intermediate states $(u(\cdot,t_{0}),m(\cdot,t_{0}))$ for any $t_{0} \in [0, T)$, i.e., the value function and the mean distribution at intermediate times, respectively. We prove conditional stability estimates under suitable assumptions on the diffusion coefficient and the initial state $(u(\cdot,0),m(\cdot,0))$. The proofs are based on Carleman's estimates with a simple weight function. We first prove a Carleman estimate for the Hamilton-Jacobi-Bellman (HJB) equation. A second Carleman estimate will be derived for the Fokker-Planck (FP) equation. Then, by combining the two estimates, we obtain a Carleman estimate for the mean-field game system, leading to the stability of the backward problems.

The phonon thermal conductance of sub-nanometric vacuum gaps between two in-plane nanoribbons of two-dimensional materials (graphene and silicene) is analyzed using the atomistic Green's function method and by employing the Tersoff and Lennard-Jones potentials for describing the interatomic interactions. It is found that the phonon conductance decays exponentially with the size of the gap. Three exponential regimes have been identified. In the regime where the Lennard-Jones (L-J) potential is driven by the repulsive interatomic forces, caused by the overlap of electronic orbits, there is a sharp exponential decay in conductance as the gap increases (exp(-10.0d) for graphene). When both the repulsive and attractive (van der Waals) interatomic forces contribute to the L-J potential, the decay rate of the conductance significantly reduces to exp(-2.0d) for graphene and exp(-2.5d) for silicene. In the regime where attractive van der Waals forces dominate the L-J potential, phonon conductance has the slowest exponential decay as exp(-1.3d) for both silicene and graphene. It is also found that the contribution from the optical phonons to the conductance is non-negligible only for very small gaps between graphene nanoribbons (d < 1.6 \AA). The phonon conductance of the gap is shown to vary with the width of the nanoribbon very modestly, such that the thermal conductivity of the gap linearly increases with the nanoribbon widths. The results of this study are of significance for fundamental understanding of heat transfer in the extreme near-field regime and for predicting the effect of interfaces and defects on heat transfer.

Accurate prediction of wind power is essential for the grid integration of this intermittent renewable source and aiding grid planners in forecasting available wind capacity. Spatial differences lead to discrepancies in climatological data distributions between two geographically dispersed regions, consequently making the prediction task more difficult. Thus, a prediction model that learns from the data of a particular climatic region can suffer from being less robust. A deep neural network (DNN) based domain adaptive approach is proposed to counter this drawback. Effective weather features from a large set of weather parameters are selected using a random forest approach. A pre-trained model from the source domain is utilized to perform the prediction task, assuming no source data is available during target domain prediction. The weights of only the last few layers of the DNN model are updated throughout the task, keeping the rest of the network unchanged, making the model faster compared to the traditional approaches. The proposed approach demonstrates higher accuracy ranging from 6.14% to even 28.44% compared to the traditional non-adaptive method.

The prediction of solar power generation is a challenging task due to its dependence on climatic characteristics that exhibit spatial and temporal variability. The performance of a prediction model may vary across different places due to changes in data distribution, resulting in a model that works well in one region but not in others. Furthermore, as a consequence of global warming, there is a notable acceleration in the alteration of weather patterns on an annual basis. This phenomenon introduces the potential for diminished efficacy of existing models, even within the same geographical region, as time progresses. In this paper, a domain adaptive deep learning-based framework is proposed to estimate solar power generation using weather features that can solve the aforementioned challenges. A feed-forward deep convolutional network model is trained for a known location dataset in a supervised manner and utilized to predict the solar power of an unknown location later. This adaptive data-driven approach exhibits notable advantages in terms of computing speed, storage efficiency, and its ability to improve outcomes in scenarios where state-of-the-art non-adaptive methods fail. Our method has shown an improvement of $10.47 \%$, $7.44 \%$, $5.11\%$ in solar power prediction accuracy compared to best performing non-adaptive method for California (CA), Florida (FL) and New York (NY), respectively.

Agriculture plays a fundamental role in driving economic growth and ensuring food security for populations around the world. Although labor-intensive agriculture has led to steady increases in food grain production in many developing countries, it is frequently challenged by adverse weather conditions, including heavy rainfall, low temperatures, and drought. These factors substantially hinder food production, posing significant risks to global food security. In order to have a profitable, sustainable, and farmer-friendly agricultural practice, this paper proposes a context-based crop recommendation system powered by a weather forecast model. For implementation purposes, we have considered the whole territory of Bangladesh. With extensive evaluation, the multivariate Stacked Bi-LSTM (three Bi-LSTM layers with a time Distributed layer) Network is employed as the weather forecasting model. The proposed weather model can forecast Rainfall, Temperature, Humidity, and Sunshine for any given location in Bangladesh with an average R-Squared value of 0.9824, and the model outperforms other state-of-the-art LSTM models. These predictions guide our system in generating viable farming decisions. Additionally, our full-fledged system is capable of alerting the farmers about extreme weather conditions so that preventive measures can be undertaken to protect the crops. Finally, the system is also adept at making knowledge-based crop suggestions for flood and drought-prone regions., Comment: 18 pages, 16 figures and 13 tables. Two figures and one table have been added to this version

Intent classification is a fundamental task in natural language understanding, aiming to categorize user queries or sentences into predefined classes to understand user intent. The most challenging aspect of this particular task lies in effectively incorporating all possible classes of intent into a dataset while ensuring adequate linguistic variation. Plenty of research has been conducted in the related domains in rich-resource languages like English. In this study, we introduce BNIntent30, a comprehensive Bengali intent classification dataset containing 30 intent classes. The dataset is excerpted and translated from the CLINIC150 dataset containing a diverse range of user intents categorized over 150 classes. Furthermore, we propose a novel approach for Bengali intent classification using Generative Adversarial BERT to evaluate the proposed dataset, which we call GAN-BnBERT. Our approach leverages the power of BERT-based contextual embeddings to capture salient linguistic features and contextual information from the text data, while the generative adversarial network (GAN) component complements the model's ability to learn diverse representations of existing intent classes through generative modeling. Our experimental results demonstrate that the GAN-BnBERT model achieves superior performance on the newly introduced BNIntent30 dataset, surpassing the existing Bi-LSTM and the stand-alone BERT-based classification model.

In this study, we looked into a number of physical properties of alkali-bismuth compounds KBi$_2$ and RbBi$_2$ in the cubic Laves phase using the density functional theory (DFT). The structural, elastic, anisotropy indices, hardness, thermo-physical parameters, electronic band structure, and optoelectronic properties have been explored. Most of the results presented in this work are novel in nature.

Federated Learning (FL) has emerged as a promising approach to address data privacy and confidentiality concerns by allowing multiple participants to construct a shared model without centralizing sensitive data. However, this decentralized paradigm introduces new security challenges, necessitating a comprehensive identification and classification of potential risks to ensure FL's security guarantees. This paper presents a comprehensive taxonomy of security and privacy challenges in Federated Learning (FL) across various machine learning models, including large language models. We specifically categorize attacks performed by the aggregator and participants, focusing on poisoning attacks, backdoor attacks, membership inference attacks, generative adversarial network (GAN) based attacks, and differential privacy attacks. Additionally, we propose new directions for future research, seeking innovative solutions to fortify FL systems against emerging security risks and uphold sensitive data confidentiality in distributed learning environments., Comment: 6 pages, 2 figures

Over the years of challenges on detecting the crash consistency of non-volatile persistent memory (PM) bugs and developing new tools to identify those bugs are quite stretching due to its inconsistent behavior on the file or storage systems. In this paper, we evaluated an open-sourced automatic bug detector tool (i.e. AGAMOTTO) to test NVM level hashing PM application to identify performance and correctness PM bugs in the persistent (main) memory. Furthermore, our faithful validation tool able to discovered 65 new NVM level hashing bugs on PMDK library and it outperformed the number of bugs (i.e. 40 bugs) that WITCHER framework was able to identified. Finally, we will propose a Deep-Q Learning search heuristic algorithm over the PM-Aware search algorithm in the state selection process to improve the searching strategy efficiently., Comment: 8 pages, 4 figures