Search and rescue (SAR) operations require optimal solutions to maximize efficiency and minimize risks for human responders. Quadruped robots have emerged as viable agents due to their locomotion and agility capabilities. This work presents a navigation framework designed to enhance stability and adaptability of legged robots in complex environments. A high-fidelity simulation environment was developed using NVIDIA IsaacSim, incorporating realistic disaster conditions such as unstructured terrain, fire, and smoke. Multiple stochastic routes were generated and analyzed in simulation in terms of energy consumption, stability, traversal time, and fall occurrences, to categorize them and determine the safest and most efficient path before real-world deployment, reducing the likelihood of failures. A fuzzy logic controller was proposed to regulate speed and improve locomotion adaptability. The proposed approach was validated in both simulation and real-world experiments. The results demonstrate the effectiveness of the proposed strategy in enabling safe and efficient navigation for quadruped robots in SAR missions. The code is publicly available at Robcib-GIT/IsaacSim_LeggedRobotSupplementary Video: https://youtu.be/cxEP5YVy8Qo

Abstract Modelling soft robots is a non‐trivial task since their behaviours rely on their morphology, materials and surrounding elements. These robots are very useful to safely interact with their environment and because of their inherent flexibility and adaptability skills. However, they are usually very hard to model because of their intrinsic non‐linearities. This fact presents a unique challenge in the computer graphics and simulation scopes. Current trends in these fields tend to narrow the gap between virtual and real environments. This work will explain a challenging modelling process for a cable‐driven soft robot called Kyma. For this purpose, the real‐time Finite Element Method (FEM) is applied using the Simulation Open Framework Architecture. And two methods are implemented and compared to optimize the model efficiency: a heuristic one and the Model Order Reduction. Both models are also validated with the real robot using a precise motion tracking system. Moreover, an analysis of robot–object interactions is proposed to test the compliance of the presented soft manipulator. As a result, the real‐time FEM emerges as a good solution to accurately and efficiently model the presented robot while also allowing to study the interactions with its environment.

Let's imagine a swarm of drones that has to visit some locations and build a map in a disaster area. Let's assume the drones only can communicate to their neighbors and manage partial information of the mission. A relevant question in this scenario is “Should the robots compete or should they cooperate?”. This work analyzes the described scenario to answer this question. Two game theoretical algorithms have been developed: one competitive and another cooperative. The competitive algorithm poses games among each drone and its neighbors and searches the Nash Equilibrium. The cooperative one defines electoral systems that allow the drones to vote their preferred task allocations for their neighbors. Both algorithms are extensively tested in multiple scenarios with different features. After the experiments the question can be answered “The robots should cooperate!”.

This paper presents a pedestrian trajectory prediction technique. Its mail novelty is that it does not require any previous observation or knowledge of pedestrian trajectories, thus making it useful for autonomous surveillance applications. The prediction requires only a set of possible goals, a map of the scenario and the initial position of the pedestrian. Then, it uses two different path planing algorithms to find the possible routes and transforms the similarity between observed and planned routes into probabilities. Finally, it applies a motion model to obtain a time-stamped predicted trajectory. The system has been used in combination with a pedestrian detection and tracking system for real-world tests as well as a simulation software for a large number of executions.

The control and monitoring of the UAV missions is a challenge in terms of operator workload. Two relevant issues are the situational awareness and the decision support of operators. This paper proposes a system that is able to analyze the telemetry of UAV to deduce the state of mission. This system uses Petri nets for determining the state and decision trees for estimating the evolution. Both the Petri nets and the decision trees are generated automatically from the telemetry of previous missions. The whole system is validated by monitoring a set of UAV missions in a realistic simulator. The results can be applied to diverse areas, including the development of intelligent and adaptive interfaces or decision support systems.

An important topic in the field of Multi Robot Systems focuses on motion coordination and synchronization for formation keeping. Although several works have addressed such problem, little attention has been devoted to study the computational complexity within the framework of large-scale systems. This paper presents our current work on how to achieve high computational performance for systems composed by a large number of robots that must fulfill with a marching and formation task. A scalable Multi-Processor Parallel Architecture is introduced with the purpose of achieving scalability, i.e., computation time of O(log2n) for a n-robots system. Our architecture has been tested onto a multi-processor system and validated against several simulations testing.

The dynamic optimization problem concerns finding an optimum in a changing environment. In the tracking problem, the optimizer should be able to follow the optimum's changes over time.In this paper we present two adaptive mutation operators designed to improve the following of a time-changing optimum, under the assumption that the changes follow a non-random law. Such law can be estimated in order to improve the optimum tracking capabilities of the algorithm. For experimental assessment, a (1,λ) evolution strategy has been applied to a dynamic version of the sphere problem.

This paper proposes the utilization of hybrid models of supervised neural networks for the modelling of dynamic systems. Particularly, as an example of a system, an autonomous helicopter or UAV is identified in both attitude and position systems. The evaluation of the model is done by comparing the radial basis and multilayer perceptron with the real system

This paper describes a mathematical model of a small-scale helicopter developed at "Universidad Politecnica of Madrid" as a part of VAMPIRA (multi-sensor aerial vehicle for autonomous inspections) project. A hybrid (analytic and empiric) model of a small helicopter is presented. The stability and response of the model is presented after a brief comparison of different techniques for helicopters dynamic modelling as well as the parameters identification using genetic algorithms. Accuracy of the model is verified by comparing real flight and simulation data obtained during the experiments.