To address the issues of low positioning accuracy and weak robustness of prior visual simultaneous localization and mapping (VSLAM) systems in dynamic environments, a semantic VSLAM (Sem-VSLAM) approach based on deep learning is proposed in this article. The proposed Sem-VSLAM algorithm adds semantic segmentation threads in parallel based on the open-source ORB-SLAM2’s visual odometry. First, while extracting the ORB features from an RGB-D image, the frame image is semantically segmented, and the segmented results are detected and repaired. Then, the feature points of dynamic objects are eliminated by using semantic information and motion consistency detection, and the poses are estimated by using the remaining feature points after the dynamic feature elimination. Finally, a 3D point cloud map is constructed by using tracking information and semantic information. The experiment uses Technical University of Munich public data to show the usefulness of the Sem-VSLAM algorithm. The experimental results show that the Sem-VSLAM algorithm can reduce the absolute trajectory error and relative attitude error of attitude estimation by about 95% compared to the ORB-SLAM2 algorithm and by about 14% compared to the VO-YOLOv5s in a highly dynamic environment and the average time consumption of tracking each frame image reaches 61 ms. It is verified that the Sem-VSLAM algorithm effectively improves the robustness and positioning accuracy in high dynamic environment and owning a satisfying real-time performance. Therefore, the Sem-VSLAM has a better mapping effect in a highly dynamic environment.

The early applications of Visual Simultaneous Localization and Mapping (VSLAM) technology were primarily focused on static environments, relying on the static nature of the environment for map construction and localization. However, in practical applications, we often encounter various dynamic environments, such as city streets, where moving objects are present. These dynamic objects can make it challenging for robots to accurately understand their own position. This paper proposes a real-time localization and mapping method tailored for dynamic environments to effectively deal with the interference of moving objects in such settings. Firstly, depth images are clustered, and they are subdivided into sub-point clouds to obtain clearer local information. Secondly, when processing regular frames, we fully exploit the structural invariance of static sub-point clouds and their relative relationships. Among these, the concept of the sub-point cloud is introduced as novel idea in this paper. By utilizing the results computed based on sub-poses, we can effectively quantify the disparities between regular frames and reference frames. This enables us to accurately detect dynamic areas within the regular frames. Furthermore, by refining the dynamic areas of keyframes using historical observation data, the robustness of the system is further enhanced. We conducted comprehensive experimental evaluations on challenging dynamic sequences from the TUM dataset and compared our approach with state-of-the-art dynamic VSLAM systems. The experimental results demonstrate that our method significantly enhances the accuracy and robustness of pose estimation. Additionally, we validated the effectiveness of the system in dynamic environments through real-world scenario tests.

Self-localization in highly dynamic environments is still a challenging problem for humanoid robots with limited computation resource. In this paper, we propose a dual-channel unscented particle filter (DC-UPF)-based localization method to address it. A key novelty of this approach is that it employs a dual-channel switch mechanism in measurement updating procedure of particle filter, solving for sparse vision feature in motion, and it leverages data from a camera, a walking odometer, and an inertial measurement unit. Extensive experiments with an NAO robot demonstrate that DC-UPF outperforms UPF and Monte–Carlo localization with regard to accuracy.

Robots are now among us and even though they compete with human beings in terms of performance and efficiency, they still fail to meet the challenge of performing a task optimally while providing strict motion safety guarantees. It is therefore necessary that the future generation of robots evolves in this direction. Generally, in robotics state-of-the-art approaches, the trajectory optimization and the motion safety issues have been addressed separately. An important contribution of this paper is to propose a motion planning method intended to simultaneously solve these two problems in a formal way. This motion planner is dubbed PassPMP-PSO. It is based on a periodic process that interleaves planning and execution for a regular update of the environment’s information. At each cycle, PassPMP-PSO computes a safe near-optimal partial trajectory using a new tree encoding technique based on particle swarm optimization (PSO). The performances of the proposed approach are firstly highlighted in simulation environments in the presence of moving objects that travel at high speed with arbitrary trajectories, while dealing with sensors field-of-view limits and occlusions. The PassPMP-PSO algorithm is tested for different tree expansions going from 13 to more than 200 nodes. The results show that for a population between 20 and 100 particles, the frequency of obtaining optimal trajectory is 100% with a rapid convergence of the algorithm to this solution. Furthermore, an experiment-based comparison demonstrates the performances of PassPMP-PSO over two other motion planning methods (the PassPMP, a previous variant of PassPMP-PSO, and the input space sampling). Finally, PassPMP-PSO algorithm is assessed through experimental tests performed on a real robotic platform using robot operating system in order to confirm simulation results and to prove its efficiency in real experiments.

A Q-learning approach is often used for navigation in static environments where state space is easy to define. In this paper, a new Q-learning approach is proposed for navigation in dynamic environments by imitating human reasoning. As a model-free method, a Q-learning method does not require the environmental model in advance. The state space and the reward function in the proposed approach are defined according to human perception and evaluation, respectively. Specifically, approximate regions instead of accurate measurements are used to define states. Moreover, due to the limitation of robot dynamics, actions for each state are calculated by introducing a dynamic window that takes robot dynamics into account. The conducted tests show that the obstacle avoidance rate of the proposed approach can reach 90.5% after training, and the robot can always operate below the dynamics limitation.

Increasingly chaotic business environments of today demand organizations to be more resilient. While the concept of resilience is widely discussed in disaster (e.g., Wildavsky, 1991) and crisis management literatures (e.g., Manyena, 2006), the literature on organizational resilience is developing disjointedly in organizational studies. The literature review suggests that some factors that are suggested in the literature as components of organizational resilience are sources contributing to the emergence of resilience in organizations. This study proposes an integrative framework for organizational resilience and introduces a new outcome concept of organizational evolvability, emphasizing the heightened sensitivity and increased wisdom of the post-event organization. In this model, sources of organizational resilience are categorized as perceptual stance, contextual integrity, strategic capacity and strategic acting, and organizational resilience leads to organizational evolvability as its outcome. The proposed organizational resilience framework attempts to provide a synthesis of the divergent literature on resilience and aims to strengthen organizational resilience research for richer theoretical and empirical progress.

The fundamentals of an approach to solving the control task of robots interaction with a dynamic environment based on the stability of a closed-loop control system are given in this paper. The task is set and solved in its general form. The traditional control concept of compliant robot motion—the hybrid position/force control is discussed. In the paper the proposed control laws ensure simultaneous stabilization of both the desired robot motion and the desired interaction force, as well as their required quality of transient responses. In order to emphasize the fundamental point of this approach in controlling the contact tasks in robotics, the authors have assumed ideal parameters of interacting dynamic systems. The proposed control procedure is demonstrated by one simple example.

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.

This work investigates the utilisation of Particle Swarm Optimisation (PSO) for the non-deterministic navigation of Unmanned Aerial Vehicles (UAVs), allowing them to work cooperatively toward the goal of protecting a wide area against airborne attack. To negate the PSO's inherent weakness in dynamic environments, a neighbourhood scheme is proposed that not only enables the efficient interception of targets several times faster than the UAVs but also facilitates the maintenance of effective airspace coverage. Empirical results suggest that these techniques may indeed be of use in autonomous navigation systems for UAVs in air defence roles.

In this paper, we present a method for robot navigation toward a moving object with unknown maneuvers. Our strategy is based on the integration of the robot and the target kinematics equations with geometric rules. The tracking problem is modeled in polar coordinates using a two-dimensional system of differential equations. The control law is then derived based on this model. Our approach consists of a rendezvous course, which means that the robot reaches the moving goal without following its path. In the presence of obstacles, two navigation modes are integrated, namely the tracking and the obstacle-avoidance modes. To confirm our theoretical results, the navigation strategy is illustrated using an extensive simulation for different scenarios.

In this research, we study the construction of configuration space (C-space) of manipulators. The proposed approach is based upon precomputing the global topology of a robot's free space, and consists of an offline phase and an online phase. In the offline phase, a C-space obstacle database (COD) for a given robot is developed in which the C-space obstacle (C-obstacle) maps are stored and indexed by the cells of the workspace; in the online phase when the same robot is operated in a real environment, those maps whose indices match the real obstacle cells are identified and then extracted from the COD. The superposition of these maps forms forbidden region in operation. This proposed approach is a generic one and can be applied to manipulators with arbitrary kinematic structures and geometries. The construction of the COD, which is generally the most time-consuming step, is implemented in the offline phase, and the online computing only involves the identification of the components matching the COD indices. Therefore, this proposed approach for C-space construction can be realized in a real-time online fashion and is especially suitable for robot manipulation under dynamic operations. We carry out analyses on several types of manipulators to verify and demonstrate the feasibility and efficiency of the proposed approach.

Based on the author's knowledge the paper gives a brief account of some of the scientific achievements of robotics that were of crucial importance to its development.

In a rough chronological order these are: zero-moment concept and semi-inverse method; recursive formulation of robot dynamics; computer-aided derivation of robot dynamics in symbolic form; dynamic approach to generation of trajectories of robotic manipulators; centralized feedforward control in robotics; robot dynamic control; decentralized control and observer applied to strongly coupled active mechanisms; force feedback in dynamic control of robots; decentralized control stability tests for robotic mechanisms; underactuated robotic systems; practical stability tests in robotics; unified approach to control laws synthesis for robot interacting with dynamic environment; modeling and control of multi-arm cooperating robots interacting with environment; connectionist algorithms for advanced learning control of robots interacting with dynamic environment; fuzzy logic robot control with model-based dynamic compensation, and internal redundancy – a new way to improve robot dynamic performance.

The paper discusses some practical problems of contact dynamics. Modelling the dynamics of contact tasks is carried out in a completely general way. Two dynamic systems, active robot system and passive environment system are brought into contact and the relevant dynamics are analyzed. The effects are: rigid-body contact force, elastodynamics in contact zone, friction in contact points, etc. Simultaneous stabilization of contact force and position is obtained using New Dynamic Position/Force Control. The general model is then applied to some more concrete problems and the simulation results are presented.

An adaptive control of multiple robot compliant manipulation of a dynamical environment is synthesized. It may be implemented when: (i) there is no good understanding of all physical effects incorporated in the multiple robots/object/environment system to be controlled; (ii) the parameters of the system are not precisely known, (iii) the system parameters do vary in a known region about their nominal values. The proposed adaptive control law has the inverse dynamics controller structure and is composed of an identification part (parameter update law), and a control law part. It is proved that the proposed adaptive controller is asymptotically stable. The simulation results verify the proposed approach to coordinated adaptive control of multiple robot manipulators in constrained motion tasks. They also verify that the multiple robots/object/environment system can track a step parameter change.

A new real-time obstacle avoidance method for mobile robots has been developed. This method, namely the vector-distance function method, permits the detection of obstacles (both moving and stationary) and generates a path that can avoid collisions. The proposed approach expresses the distance information in a vector form. Then the notion of weighting is introduced to describe relationship between sensors of mobile robots and the target to be reached. Furthermore, R-mode, L-mode and T-mode are introduced to generate a safe path for the mobile robot in a dynamic environment filled with both stationary and moving obstacles. The algorithm can deal with a complicated obstacle environment, such as multiple concave and convex obstacles. Simulation results are included to demonstrate the applicability and effectiveness of the developed algorithm.