arXiv subjects

Panayiotis Kolios

Publications and source records attributed to Panayiotis Kolios.

At least 37 records · Page 2Linked to original sources

arXiv2024

Distributed Control for 3D Inspection using Multi-UAV Systems

Cooperative control of multi-UAV systems has attracted substantial research attention due to its significance in various application sectors such as emergency response, search and rescue missions, and critical infrastructure inspection. This paper proposes a distributed control algorithm to generate collision-free trajectories that drive the multi-UAV system to completely inspect a set of 3D points on the surface of an object of interest. The objective of the UAVs is to cooperatively inspect the object of interest in the minimum amount of time. Extensive numerical simulations for a team of quadrotor UAVs inspecting a real 3D structure illustrate the validity and effectiveness of the proposed approach.

cs.RO↗

arXiv2024

Model Predictive Control For Multiple Castaway Tracking with an Autonomous Aerial Agent

Over the past few years, a plethora of advancements in Unmanned Areal Vehicle (UAV) technology has paved the way for UAV-based search and rescue operations with transformative impact to the outcome of critical life-saving missions. This paper dives into the challenging task of multiple castaway tracking using an autonomous UAV agent. Leveraging on the computing power of the modern embedded devices, we propose a Model Predictive Control (MPC) framework for tracking multiple castaways assumed to drift afloat in the aftermath of a maritime accident. We consider a stationary radar sensor that is responsible for signaling the search mission by providing noisy measurements of each castaway's initial state. The UAV agent aims at detecting and tracking the moving targets with its equipped onboard camera sensor that has limited sensing range. In this work, we also experimentally determine the probability of target detection from real-world data by training and evaluating various Convolutional Neural Networks (CNNs). Extensive qualitative and quantitative evaluations demonstrate the performance of the proposed approach.

cs.RO↗

arXiv2024

Probabilistically Robust Trajectory Planning of Multiple Aerial Agents

Current research on robust trajectory planning for autonomous agents aims to mitigate uncertainties arising from disturbances and modeling errors while ensuring guaranteed safety. Existing methods primarily utilize stochastic optimal control techniques with chance constraints to maintain a minimum distance among agents with a guaranteed probability. However, these approaches face challenges, such as the use of simplifying assumptions that result in linear system models or Gaussian disturbances, which limit their practicality in complex realistic scenarios. To address these limitations, this work introduces a novel probabilistically robust distributed controller enabling autonomous agents to plan safe trajectories, even under non-Gaussian uncertainty and nonlinear systems. Leveraging exact uncertainty propagation techniques based on mixed-trigonometric-polynomial moment propagation, this method transforms non-Gaussian chance constraints into deterministic ones, seamlessly integrating them into a distributed model predictive control framework solvable with standard optimization tools. Simulation results demonstrate the effectiveness of this technique, highlighting its ability to consistently handle various types of uncertainty, ensuring robust and accurate path planning in complex scenarios.

eess.SY↗

arXiv2024

Adaptive Autopilot: Constrained DRL for Diverse Driving Behaviors

In pursuit of autonomous vehicles, achieving human-like driving behavior is vital. This study introduces adaptive autopilot (AA), a unique framework utilizing constrained-deep reinforcement learning (C-DRL). AA aims to safely emulate human driving to reduce the necessity for driver intervention. Focusing on the car-following scenario, the process involves (i) extracting data from the highD natural driving study and categorizing it into three driving styles using a rule-based classifier; (ii) employing deep neural network (DNN) regressors to predict human-like acceleration across styles; and (iii) using C-DRL, specifically the soft actor-critic Lagrangian technique, to learn human-like safe driving policies. Results indicate effectiveness in each step, with the rule-based classifier distinguishing driving styles, the regressor model accurately predicting acceleration, outperforming traditional car-following models, and C-DRL agents learning optimal policies for humanlike driving across styles.

cs.RO↗

arXiv2024

Edge-Assisted ML-Aided Uncertainty-Aware Vehicle Collision Avoidance at Urban Intersections

Intersection crossing represents one of the most dangerous sections of the road infrastructure and Connected Vehicles (CVs) can serve as a revolutionary solution to the problem. In this work, we present a novel framework that detects preemptively collisions at urban crossroads, exploiting the Multi-access Edge Computing (MEC) platform of 5G networks. At the MEC, an Intersection Manager (IM) collects information from both vehicles and the road infrastructure to create a holistic view of the area of interest. Based on the historical data collected, the IM leverages the capabilities of an encoder-decoder recurrent neural network to predict, with high accuracy, the future vehicles' trajectories. As, however, accuracy is not a sufficient measure of how much we can trust a model, trajectory predictions are additionally associated with a measure of uncertainty towards confident collision forecasting and avoidance. Hence, contrary to any other approach in the state of the art, an uncertainty-aware collision prediction framework is developed that is shown to detect well in advance (and with high reliability) if two vehicles are on a collision course. Subsequently, collision detection triggers a number of alarms that signal the colliding vehicles to brake. Under real-world settings, thanks to the preemptive capabilities of the proposed approach, all the simulated imminent dangers are averted.

cs.LG↗

arXiv2024

Automated Real-Time Inspection in Indoor and Outdoor 3D Environments with Cooperative Aerial Robots

This work introduces a cooperative inspection system designed to efficiently control and coordinate a team of distributed heterogeneous UAV agents for the inspection of 3D structures in cluttered, unknown spaces. Our proposed approach employs a two-stage innovative methodology. Initially, it leverages the complementary sensing capabilities of the robots to cooperatively map the unknown environment. It then generates optimized, collision-free inspection paths, thereby ensuring comprehensive coverage of the structure's surface area. The effectiveness of our system is demonstrated through qualitative and quantitative results from extensive Gazebo-based simulations that closely replicate real-world inspection scenarios, highlighting its ability to thoroughly inspect real-world-like 3D structures.

cs.RO↗

arXiv2024

Hierarchical Fault-Tolerant Coverage Control for an Autonomous Aerial Agent

Fault-tolerant coverage control involves determining a trajectory that enables an autonomous agent to cover specific points of interest, even in the presence of actuation and/or sensing faults. In this work, the agent encounters control inputs that are erroneous; specifically, its nominal controls inputs are perturbed by stochastic disturbances, potentially disrupting its intended operation. Existing techniques have focused on deterministically bounded disturbances or relied on the assumption of Gaussian disturbances, whereas non-Gaussian disturbances have been primarily been tackled via scenario-based stochastic control methods. However, the assumption of Gaussian disturbances is generally limited to linear systems, and scenario-based methods can become computationally prohibitive. To address these limitations, we propose a hierarchical coverage controller that integrates mixed-trigonometric-polynomial moment propagation to propagate non-Gaussian disturbances through the agent's nonlinear dynamics. Specifically, the first stage generates an ideal reference plan by optimising the agent's mobility and camera control inputs. The second-stage fault-tolerant controller then aims to follow this reference plan, even in the presence of erroneous control inputs caused by non-Gaussian disturbances. This is achieved by imposing a set of deterministic constraints on the moments of the system's uncertain states.

eess.SY↗

arXiv2024

Cooperative Receding Horizon 3D Coverage Control with a Team of Networked Aerial Agents

This work proposes a receding horizon coverage control approach which allows multiple autonomous aerial agents to work cooperatively in order cover the total surface area of a 3D object of interest. The cooperative coverage problem which is posed in this work as an optimal control problem, jointly optimizes the agents' kinematic and camera control inputs, while considering coupling constraints amongst the team of agents which aim at minimizing the duplication of work. To generate look-ahead coverage trajectories over a finite planning horizon, the proposed approach integrates visibility constraints into the proposed coverage controller in order to determine the visible part of the object with respect to the agents' future states. In particular, we show how non-linear and non-convex visibility determination constraints can be transformed into logical constraints which can easily be embedded into a mixed integer optimization program.

cs.RO↗

arXiv2024

Pandemic infection forecasting through compartmental model and learning-based approaches

The emergence and spread of deadly pandemics has repeatedly occurred throughout history, causing widespread infections and loss of life. The rapid spread of pandemics have made governments across the world adopt a range of actions, including non-pharmaceutical measures to contain its impact. However, the dynamic nature of pandemics makes selecting intervention strategies challenging. Hence, the development of suitable monitoring and forecasting tools for tracking infected cases is crucial for designing and implementing effective measures. Motivated by this, we present a hybrid pandemic infection forecasting methodology that integrates compartmental model and learning-based approaches. In particular, we develop a compartmental model that includes time-varying infection rates, which are the key parameters that determine the pandemic's evolution. To identify the time-dependent infection rates, we establish a hybrid methodology that combines the developed compartmental model and tools from optimization and neural networks. Specifically, the proposed methodology estimates the infection rates by fitting the model to available data, regarding the COVID-19 pandemic in Cyprus, and then predicting their future values through either a) extrapolation, or b) feeding them to neural networks. The developed approach exhibits strong accuracy in predicting infections seven days in advance, achieving low average percentage errors both using the extrapolation (9.90%) and neural network (5.04%) approaches.

q-bio.PE↗

arXiv2023

A Study of Data-driven Methods for Adaptive Forecasting of COVID-19 Cases

Severe acute respiratory disease SARS-CoV-2 has had a found impact on public health systems and healthcare emergency response especially with respect to making decisions on the most effective measures to be taken at any given time. As demonstrated throughout the last three years with COVID-19, the prediction of the number of positive cases can be an effective way to facilitate decision-making. However, the limited availability of data and the highly dynamic and uncertain nature of the virus transmissibility makes this task very challenging. Aiming at investigating these challenges and in order to address this problem, this work studies data-driven (learning, statistical) methods for incrementally training models to adapt to these nonstationary conditions. An extensive empirical study is conducted to examine various characteristics, such as, performance analysis on a per virus wave basis, feature extraction, "lookback" window size, memory size, all for next-, 7-, and 14-day forecasting tasks. We demonstrate that the incremental learning framework can successfully address the aforementioned challenges and perform well during outbreaks, providing accurate predictions.

cs.LG↗

arXiv2023

A Multi-task Learning Framework for Drone State Identification and Trajectory Prediction

The rise of unmanned aerial vehicle (UAV) operations, as well as the vulnerability of the UAVs' sensors, has led to the need for proper monitoring systems for detecting any abnormal behavior of the UAV. This work addresses this problem by proposing an innovative multi-task learning framework (MLF-ST) for UAV state identification and trajectory prediction, that aims to optimize the performance of both tasks simultaneously. A deep neural network with shared layers to extract features from the input data is employed, utilizing drone sensor measurements and historical trajectory information. Moreover, a novel loss function is proposed that combines the two objectives, encouraging the network to jointly learn the features that are most useful for both tasks. The proposed MLF-ST framework is evaluated on a large dataset of UAV flights, illustrating that it is able to outperform various state-of-the-art baseline techniques in terms of both state identification and trajectory prediction. The evaluation of the proposed framework, using real-world data, demonstrates that it can enable applications such as UAV-based surveillance and monitoring, while also improving the safety and efficiency of UAV operations.

eess.SY↗

arXiv2023

Joint Estimation and Control for Multi-Target Passive Monitoring with an Autonomous UAV Agent

This work considers the problem of passively monitoring multiple moving targets with a single unmanned aerial vehicle (UAV) agent equipped with a direction-finding radar. This is in general a challenging problem due to the unobservability of the target states, and the highly non-linear measurement process. In addition to these challenges, in this work we also consider: a) environments with multiple obstacles where the targets need to be tracked as they manoeuvre through the obstacles, and b) multiple false-alarm measurements caused by the cluttered environment. To address these challenges we first design a model predictive guidance controller which is used to plan hypothetical target trajectories over a rolling finite planning horizon. We then formulate a joint estimation and control problem where the trajectory of the UAV agent is optimized to achieve optimal multi-target monitoring.

eess.SY↗

arXiv2023

Unscented Optimal Control for 3D Coverage Planning with an Autonomous UAV Agent

We propose a novel probabilistically robust controller for the guidance of an unmanned aerial vehicle (UAV) in coverage planning missions, which can simultaneously optimize both the UAV's motion, and camera control inputs for the 3D coverage of a given object of interest. Specifically, the coverage planning problem is formulated in this work as an optimal control problem with logical constraints to enable the UAV agent to jointly: a) select a series of discrete camera field-of-view states which satisfy a set of coverage constraints, and b) optimize its motion control inputs according to a specified mission objective. We show how this hybrid optimal control problem can be solved with standard optimization tools by converting the logical expressions in the constraints into equality/inequality constraints involving only continuous variables. Finally, probabilistic robustness is achieved by integrating the unscented transformation to the proposed controller, thus enabling the design of robust open-loop coverage plans which take into account the future posterior distribution of the UAV's state inside the planning horizon.

cs.RO↗

arXiv2023

Transitioning towards fit-for-purpose Public Health Surveillance Systems

The COVID-19 pandemic has exposed several weaknesses in the public health infrastructure, including supply chain mechanisms and public health ICT systems. The expansion of testing and contact tracing has been key to identifying and isolating infected individuals, as well as tracking and containing the spread of the virus. Digital technologies, such as telemedicine and virtual consultations, have experienced a surge in demand to provide medical support while minimizing the risk of transmission and infection. The pandemic has made it clear that cooperation, information sharing, and communication among stakeholders are crucial in making the right decisions and preventing future outbreaks. Redesigning public health systems for effective management of outbreaks should include five key elements: disease surveillance and early warning systems, contact tracing and case management, data analytics and visualization, communication and education, and telemedicine. As the world navigates the COVID-19 pandemic, healthcare ICT systems will play an increasingly important role in the future of healthcare delivery. In a post COVID-19 world, several ICT strategies should be implemented to improve the quality, efficiency, and accessibility of healthcare services, including the expansion of telemedicine, data analytics and population health management, interoperability, and cybersecurity. Overall, this report summarises the importance of early detection and rapid response, international cooperation and coordination, clear and consistent communication, investing in public health systems and emergency preparedness, digital technology and telemedicine, and equity and social determinants of health. These lessons demonstrate the need for better preparedness and planning for future crises and the importance of addressing underlying issues to create a more resilient and accessible digital infrastructure.

cs.CY↗

arXiv2023

Distributed Estimation and Control for Jamming an Aerial Target With Multiple Agents

This work proposes a distributed estimation and control approach in which a team of aerial agents equipped with radio jamming devices collaborate in order to intercept and concurrently track-and-jam a malicious target, while at the same time minimizing the induced jamming interference amongst the team. Specifically, it is assumed that the malicious target maneuvers in 3D space, avoiding collisions with obstacles and other 3D structures in its way, according to a stochastic dynamical model. Based on this, a track-and-jam control approach is proposed which allows a team of distributed aerial agents to decide their control actions online, over a finite planning horizon, to achieve uninterrupted radio-jamming and tracking of the malicious target, in the presence of jamming interference constraints. The proposed approach is formulated as a distributed model predictive control (MPC) problem and is solved using mixed integer quadratic programming (MIQP). Extensive evaluation of the system's performance validates the applicability of the proposed approach in challenging scenarios with uncertain target dynamics, noisy measurements, and in the presence of obstacles.

eess.SY↗

arXiv2023

Integrated Guidance and Gimbal Control for Coverage Planning With Visibility Constraints

Coverage path planning with unmanned aerial vehicles (UAVs) is a core task for many services and applications including search and rescue, precision agriculture, infrastructure inspection and surveillance. This work proposes an integrated guidance and gimbal control coverage path planning (CPP) approach, in which the mobility and gimbal inputs of an autonomous UAV agent are jointly controlled and optimized to achieve full coverage of a given object of interest, according to a specified set of optimality criteria. The proposed approach uses a set of visibility constraints to integrate the physical behavior of sensor signals (i.e., camera-rays) into the coverage planning process, thus generating optimized coverage trajectories that take into account which parts of the scene are visible through the agent's camera at any point in time. The integrated guidance and gimbal control CPP problem is posed in this work as a constrained optimal control problem which is then solved using mixed integer programming (MIP) optimization. Extensive numerical experiments demonstrate the effectiveness of the proposed approach.

eess.SY↗

arXiv2023

UAV-based Receding Horizon Control for 3D Inspection Planning

Nowadays, unmanned aerial vehicles or UAVs are being used for a wide range of tasks, including infrastructure inspection, automated monitoring and coverage. This paper investigates the problem of 3D inspection planning with an autonomous UAV agent which is subject to dynamical and sensing constraints. We propose a receding horizon 3D inspection planning control approach for generating optimal trajectories which enable an autonomous UAV agent to inspect a finite number of feature-points scattered on the surface of a cuboid-like structure of interest. The inspection planning problem is formulated as a constrained open-loop optimal control problem and is solved using mixed integer programming (MIP) optimization. Quantitative and qualitative evaluation demonstrates the effectiveness of the proposed approach.

cs.RO↗

arXiv2023

Integrated Ray-Tracing and Coverage Planning Control using Reinforcement Learning

In this work we propose a coverage planning control approach which allows a mobile agent, equipped with a controllable sensor (i.e., a camera) with limited sensing domain (i.e., finite sensing range and angle of view), to cover the surface area of an object of interest. The proposed approach integrates ray-tracing into the coverage planning process, thus allowing the agent to identify which parts of the scene are visible at any point in time. The problem of integrated ray-tracing and coverage planning control is first formulated as a constrained optimal control problem (OCP), which aims at determining the agent's optimal control inputs over a finite planning horizon, that minimize the coverage time. Efficiently solving the resulting OCP is however very challenging due to non-convex and non-linear visibility constraints. To overcome this limitation, the problem is converted into a Markov decision process (MDP) which is then solved using reinforcement learning. In particular, we show that a controller which follows an optimal control law can be learned using off-policy temporal-difference control (i.e., Q-learning). Extensive numerical experiments demonstrate the effectiveness of the proposed approach for various configurations of the agent and the object of interest.

cs.RO↗