Motion planning for autonomous vehicles requires generating collision-free and dynamically feasible trajectories in complex environments under real-time constraints. While nonlinear optimal control formulations provide high-fidelity solutions, they are computationally demanding and sensitive to initialization, whereas geometric planning methods scale well but often decouple path selection from trajectory optimization. This paper studies the extent to which optimization over Graphs of Convex Sets (GCS) can approximate solutions of nonlinear optimal control problems in the context of autonomous driving. The free space is represented as a finite union of convex regions organized as a directed graph, allowing nonconvex geometry to be handled through discrete connectivity decisions while maintaining convex trajectory constraints within each region. Vehicle motion is parameterized using Bezier curves for the spatial path and a polynomial time-scaling function for temporal evolution. Under small-slip and linear tire assumptions, a simplified dynamic bicycle model enables approximate enforcement of dynamic feasibility through convex constraints on trajectory derivatives. The approach is evaluated in CommonRoad scenarios involving static obstacle avoidance and lane-changing maneuvers, and is compared against a nonlinear discrete-time optimal control formulation. The results indicate that the GCS-based method generates collision-free and dynamically consistent trajectories that closely match those obtained from the nonlinear program, while exhibiting improved computational efficiency and reduced sensitivity to initialization. These findings suggest that GCS provides a structured approximation of nonlinear motion planning problems, capturing dominant geometric and dynamic effects while preserving convexity in the continuous relaxation.

The stochastic nature of time delays in networked control systems poses significant challenges for controller synthesis and the corresponding analyses, leading to conservative designs and degraded performance. Existing approaches approximate stochastic delays by fixed, worst-case values, which limits their ability to describe the behavior of distributed control implementations, including distributed computational architectures, multi-core processing, and network communication between different computational nodes. Thus, with regard to this context, this work proposes a novel modeling framework for linear multiple-input multiple-output networked control systems that represents stochastic sampling instants and delays through a stochastic linear time-varying state-space model. Based on this model, a predictor-based compensator derived from the filtered Smith predictor is developed to mitigate the effects of stochastic time delays. Using a cooperative adaptive cruise control benchmark, the proposed compensator is compared with a baseline fixed-delay predictor in terms of performance degradation, with all metrics normalized relative to the delay-free case. Numerical results indicate that the proposed method achieves a 55% reduction in the worst-case tracking error energy with respect to the baseline controller, and a 65% reduction in the worst-case control effort.

This study employs physics-informed neural ordinary differential equations to perform time series forecasting for dynamic system monitoring. In this approach, predictions of the system’s dynamic response over time are generated by integrating physical laws with observed data, enhancing forecasting accuracy. The dataset consists of vibration data from radial and tangential accelerometers, recorded over a 5-second interval under various conditions such as faults, imbalances, and misalignments. The results reveal that the fully informed forecasting model consistently produces lower kurtosis values and a more normally distributed error, suggesting superior model generalization over time. Meanwhile, the partially informed model also shows improved performance, with fewer biased predictions than the uninformed model, which exhibits higher kurtosis and a greater number of outliers in its forecasts. Importantly, the study demonstrates that the model can successfully extrapolate system behavior beyond the training period, maintaining high prediction accuracy despite the limited amount of data. Overall, these findings underscore the potential of using Physics-Informed Neural Ordinary Differential Equations for reliable time series forecasting in dynamic system monitoring.

The exponential growth of data in astrophysics, seismology, and climate modeling demands efficient compression strategies to optimize storage, transmission, and computational resources. This paper presents a modular and extensible compression framework tailored for big data applications, enabling dynamic reconfiguration of compression pipelines based on contextual requirements. The proposed framework allows for the integration of domain -specific algorithms aiming at achieving greater compression ratios by selecting compression strategies that are more suitable for a given dataset. The proposed framework was evaluated through a comprehensive benchmarking process and compared with two traditional data compression tools, FPZIP and ZFP, regarding both the compression ratio and computational performance. The results highlight the advantages of adaptive, fine-tuned compression strategies in managing largescale datasets efficiently, as the proposed approach achieves greater compression ratios due to its flexibility in selecting compression strategies appropriate for a given domain.

Industrial application data acquisition systems can be sources of vast amounts of data. The seismic surveys conducted by oil and gas companies result in enormous datasets, often exceeding terabytes of data. The storage and communication demands these data require can only be achieved through compression. Careful consideration must be given to minimize the reconstruction error of compressed data caused by lossy compression. This paper investigates the combination of principal component analysis (PCA), discrete wavelet transform (DWT), thresholding, quantization, and entropy encoding to compress such datasets. The proposed method is a lossy compression algorithm tuned by evaluating the reconstruction error in frequency ranges of interest, namely 0–20 Hz and 15–65 Hz. The PCA compression and decompression acts as a noise filter while the DWT drives the compression. The proposed method can be tuned through threshold and quantization percentages and the number of principal components to achieve compression rates of up to 31:1 with reconstruction residues energy of less than 4% in the frequency ranges of 0–20 Hz, 15–65 Hz, and 60–105 Hz.

The seismic surveys conducted by the oil and gas sector result in very large datasets, often exceeding terabytes of data, leading to high costs and technical challenges regarding storage and transmission of such large quantities of data. Therefore, data compression is crucial to address the challenges related to communication and storage demands. This paper investigates a compression strategy based on the combined application of the Discrete Cosine Transform and Principal Component Analysis and its ability to achieve higher compression ratios than the application of each of those methods alone. A theoretical motivation for the increased compression performance is presented, emphasizing that the application of Principal Component Analysis to signals in the transformed domain make the resulting signal more suitable for the posterior Thresholding, Quantization and Entropy Encoding steps. The proposed method was evaluated using a dataset containing passive seismic data collected during an oil and gas reservoir monitoring survey, achieving compression ratios of up to 1000:1 with a normalized reconstruction residue energy of less than 20%, relative to the original signal energy, for the more relevant frequency range between 0 and 20 Hz, outperforming other seismic data compression strategies considered in the literature.

This work addresses the use of Physics-Informed Neural Ordinary Differential Equations (PINODEs) for the identification and monitoring of dynamic systems. By combining prior knowledge regarding the physics that governs the behavior of a dynamic system with the flexibility and learning capabilities of Artificial Neural Networks (ANNs), it is possible employ data-driven methods that result in a robust representation of the systems, even without full knowledge of the underlying physical phenomena, while retaining a degree of interpretability of the model’s outputs. A description of the overall framework for modeling the system identification problem and training the ANNs under the is presented, along with an application case study for condition monitoring of a wind turbine’s gearbox using vibration data. The results demonstrate the ability of the identified model to generalize with great accuracy to scenarios not accounted for during training, a property attributed to the inclusion of information regarding the physics of the problem during the training procedure. It is also shown that when exposed to data collected during a fault condition, the model’s output significantly deviate from the actual measurements, hence its potential use as a tool for condition monitoring and fault detection is also successfully demonstrated.

This work addresses the use of a machine-learning based confidence attribution scheme to detect GPS spoofing attacks against cyber-physical systems. The confidence attribution scheme assigns a continuous value, between 0% and 100%, to each measurement, expressing how consistent this measurement is with measurements from other sensors in a cyber-physical system. By verifying the consistency of GPS measurements when compared to other sensors available, such as those present in an inertial measurement unit, it is possible to detect GPS spoofing attacks. The solution is evaluated on an experimentally acquired dataset containing flight logs, including sensor measurements and control signals, of an UAV subject to GPS spoofing attacks. The results show the number of anomalies detected by the confidence attribution scheme increases significantly in the presence of a GPS spoofing attack, rendering its detection.

Cyber-Physical Systems (CPS) rely on sensing to control and optimize their operation. Nevertheless, sensing itself is prone to errors that can originate at several stages, from sampling to communication. In this context, several systems adopt multivariate predictors to assess the quality of the sensed data, to replace data from faulty sensors, or to derive variables that cannot be directly sensed. These predictors are often evaluated based on their accuracy and computing demands, however, such evaluations often do not consider the system's architecture from a broader perspective, ignoring the way components are interconnected and how they cascade as inputs of other Machine Learning (ML) models. In this work, we introduce a method to evaluate the performance of interdependent predictors based on the stability of the estimation error dynamics in faulty scenarios. The proposed method estimates the ability of a predictor to produce accurate predictions while accounting for the impacts of cascading predicted values as its inputs. The prediction correctness is estimated based solely on information acquired during the training of the multivariate predictors and mathematical properties of the ML activation functions. The proposed method is evaluated with a meaningful dataset in the scope of monitoring and control of a Cyber-Physical System, and the evaluation demonstrates the ability of the proposed method to account for the interdependence of data predictors.

This paper presents a review of formal methods, covering both timed automata and Signal Temporal Logic (STL) approaches, and proposes an integration of formal methods with a data-driven representation of an Autonomous Vehicles (AV) case study. The data-driven representation of the system is done through the concept of SmartData, a data construct that includes concepts of location, timing, and semantics, providing an alternative to represent critical systems through the data they rely on. The timing and dependency relationship between different SmartData are derived into an STL expression that specifies the property monitors to verify each piece of data. The same verification is also presented in the form of timed automata, a closer representation of the tools adopted for runtime verification. The SmartData representation and STL and timed automata models are depicted through a case study considering an autonomous vehicles application. Finally, we demonstrate a general scenario for mapping data-driven systems using SmartData directly into timed automata.

In this paper, we use multivariate machine learning-based predictors to replace missing data and propose a mechanism to evaluate and track correctness by estimating its confidence level whenever successive missing data points occur. The proposed solution relies on the idea of confidence attribution, which assigns a value to every measurement, indicating how much it is believed to be accurate based on the difference between measured and predicted data. When data is missing, we perform data imputation using the predicted value and estimate confidence. We estimate confidence based solely on parameters used for confidence attribution and information acquired during the predictor's training. We evaluate the solution with two real datasets, one collected from a solar farm and another from a collection of wind turbines. The results show that the accuracy of multivariate models can decrease significantly when input data goes missing, demonstrating the need for the proposed confidence tracking mechanism.