Elias K. Xidias

Mission design for a group of autonomous guided vehicles

In this paper, a generic approach for the integration of vehicle routing and scheduling and motion planning for a group of autonomous guided vehicles (AGVs) is proposed. The AGVs are requested to serve all the work stations in a two-dimensional environment while taking into account kinematics constraints and the environments geometry during their motion. The problem objective is the simultaneous determination of time-optimum and collision-free paths for all AGVs. The proposed method is investigated and discussed through a number of simulated experiments using a variety of environments and different initial conditions.

Time-optimal task scheduling for articulated manipulators in environments cluttered with obstacles

This paper proposes a new approach for solving a generalization of the task scheduling problem for articulated robots (either redundant or non-redundant), where the robots 2D environment is cluttered with obstacles of arbitrary size, shape and location, while a set of task-points are located in the robots free-space. The objective is to determine the optimum collision-free robots tip tour through all task-points passing from each one exactly once and returning to the initial task-point. This scheduling problem combines two computationally NP-hard problems: the optimal scheduling of robot tasks and the collision-free motion planning between the task-points. The proposed approach employs the bump-surface (B-Surface) concept for the representation of the 2D robots environment by a B-Spline surface embedded in 3D Euclidean space. The time-optimal task scheduling is being searched on the generated B-Surface using a genetic algorithm (GA) with a special encoding in order to take into consideration the infinite configurations corresponding to each task-point. The result of the GAs searching constitutes the solution to the task scheduling problem and satisfies optimally the task scheduling criteria and objectives. Extensive experimental results show the efficiency and the effectiveness of the proposed method to determine the collision-free motion among obstacles.

A cloud based architecture capable of perceiving and predicting multiple vessel behaviour

We train an Artificial Neural Network (ANN) to predict future vessels behavior.We study if an ANN is capable of inferring the unique behaviour of a vessel.We design, train and implement a proof of concept ANN as a cloud based web app.The derived ANN has the ability to predict short and long term vessel behaviour. Progressively huge amounts of data, tracking vessels during their voyages across the seas, are becoming available, mostly due to the automatic identification system (AIS) that vessels of specific categories are required to carry. These datasets provide detailed insights into the patterns vessels follow, while safely navigating across the globe, under various conditions. In this paper, we develop an Artificial Neural Network (ANN) capable of predicting a vessels future behaviour (position, speed and course), based on events that occur in a predictable pattern, across large map areas. The main concept of this study is to determine if an ANN is capable of inferring the unique behavioural patterns that each vessel follows and successively use this as a means for predicting multiple vessel behaviour into a future point in time. We design, train and implement a proof of concept ANN, as a cloud based web application, with the ability of overlaying predicted short and long term vessel behaviour on an interactive map. Our proposed approach could potentially assist in busy port scheduling, vessel route planning, anomaly detection and increasing overall Maritime Domain Awareness.

Motion planning for multiple non-holonomic robots: A geometric approach

In this paper, a geometrical approach is developed to generate simultaneously optimal (or near-optimal) smooth paths for a set of non-holonomic robots, moving only forward in a 2D environment cluttered with static and moving obstacles. The robots environment is represented by a 3D geometric entity called Bump-Surface, which is embedded in a 4D Euclidean space. The multi-motion planning problem (MMPP) is resolved by simultaneously finding the paths for the set of robots represented by monoparametric smooth C2 curves onto the Bump-Surface, such that their inverse images onto the initial 2D workspace satisfy the optimization motion-planning criteria and constraints. The MMPP is expressed as an optimization problem, which is solved on the Bump-Surface using a genetic algorithm. The performance of the proposed approach is tested through a considerable number of simulated 2D dynamic environments with car-like robots.

Two-dimensional motion-planning for nonholonomic robots using the bump-surfaces concept

In this paper, a new method is introduced for finding a near-optimal path of a nonholonomic robot moving in a 2D environment cluttered with static obstacles. The method is based on the Bump-Surfaces concept and is able to deal with robots represented by a translating and rotating rigid body. The proposed approach is applied to car-like robots.

Vehicle scheduling in 2D shop floor environments

Purpose – The purpose of this paper is to develop an efficient method for solving a vehicle scheduling problem (VSP) in 2D industrial environments. An autonomous vehicle is requested to serve a set of work centers in the shop floor providing transport and delivery tasks while avoiding collisions with obstacles during its travel. The objective is to find a minimum in length, collision‐free vehicle routing schedule that serves timely as many as possible work centers in the shop floor.Design/methodology/approach – First, the vehicles environment is mapped into a 2D B‐Spline surface embedded in 3D Euclidean space using a robust geometric model. Then, a modified genetic algorithm is applied on the generated surface to search for an optimum legal route schedule that satisfies the requirements of the vehicles mission.Findings – Simulation experiments show that the method is robust enough and can determine in a reasonable computation time a solution to VSP under consideration.Originality/value – There is a gap ...

Real-time vessel behavior prediction

Abstract Vessel traffic management systems (VTMS) and vessel traffic monitoring information systems (VTMIS) have been available for a number of years now. These systems have significantly contributed to increasing the efficiency and safety of operations at sea. However, nowadays, risks at sea are once again on the rise, thus demanding an evolution in VTMS and VTMIS, such that they can support a human operator’s better understanding of the complex reality at sea and enhance his or her decision-making in light of danger. A critical requirement of such systems, is that they exhibit the ability to for-see unfolding cautious and potentially hazardous situations, so as to propose measures of danger avoidance. In this study, we employ machine learning, and specifically artificial neural networks, as a tool to add predictive capacity to VTMIS. The main objective of this study is to implement a publicly accessible, web-based system capable of real time learning and accurately predicting any vessels future behavior in low computational time. This work describes our approach, design choices, implementation and evaluation details, while we present a proof of concept prototype system. Our proposal can potentially be used as the predictive foundation for various intelligent systems, including vessel collision prevention, vessel route planning, operation efficiency estimation and even anomaly detection systems.

Energy-Minimizing Motion Design for Nonholonomic Robots Amidst Moving Obstacles

AbstractIn this paper, a new method is introduced for computing energy-minimizing motions in twodimensional environments cluttered with a priori known static and moving obstacles. The proposed method is based on a new four-dimensional motion-planning space represented by a Bump- Surface embedded in ℜ4. The energy-minimizing motion-design problem is expressed by a variational curve-design problem on the introduced Bump-Surface. The optimal motion is determined by a new algorithm for computing a B-spline curve on the Bump-Surface which is conformal to the motion-planning constrains. The proposed method is applied for designing the motion of nonholonomic robots in the presence of moving obstacles and its performance is tested in simulated 2D dynamic environments with car-like robots.

Path Planning and scheduling for a fleet of autonomous vehicles

This paper presents a new solution approach for managing the motion of a fleet of autonomous vehicles (AVs) in indoor factory environments. AVs are requested to serve a number of workstations (WS) (following a specified desired production plan for materials requirements) while taking into account the safe movement (collisions avoidance) in the shop floor as well as time duration and energy resources. The proposed approach is based on the Bump-Surface concept to represent the 2D environment through a single mathematical entity. The solution of the combined problem of path planning and task scheduling is searched on a higher-dimension B-surface (in our case 3D) in such a way that its inverse image into the robot environment satisfies the given objectives and constraints. Then, a modified Genetic Algorithm (GA) is used to search for a near-optimum solution. The objective of the fleet coordination consists of determining the best feasible paths for the AVs so that all the WS are served at the lowest possible cost. The efficiency of the developed method is investigated and discussed through characteristic simulated experiments concerning a variety of operating environments.

Continuous curvature constrained shortest path for a car-like robot using S-Roadmaps

This paper proposes a new approach for motion planning for a car-like robot which is based on a surfaced roadmap concept. A major advantage of our approach is that, it enables the same roadmap to be efficiently utilized for car-like robots with different kinematical constraints or with different starting and ending points. Our approach first represents the 2D environment using the B-Surface concept. Then, build a roadmap onto the B-Surface which does not incorporate any kinematical constraints. The paths encoded in the roadmap consist of poly-geodesic segments. The roadmap assists in the optimization and smoothing of these paths using NURBS curves. We also demonstrate experimental results for a simple model of a car-like robot moving on 2D environments.