BEASST: Behavioral Entropic Gradient based Adaptive Source Seeking for Mobile Robots

Memory loss and the inability to form new memories properly is a prominent and devastating feature of many widespread neurological conditions, particularly Alzheimer's disease and other forms of dementia. Memory function is supported by the limbic system, and more specifically the ability to remember recent events requires the hippocampus and associated structures including the entorhinal, perirhinal, and parahippocampal cortices. In rodent models, scientists have found that direct electrical stimulation of the perforant pathway, which arises from the entorhinal cortex in the medial temporal lobe and projects to the hippocampus, results in the cellular correlates of long term potentiation. In a new study, Suthana et al (Memory enhancement and deep-brain stimulation of the entorhinal area. N Engl J Med. 2012;366[6]:502-510.) have demonstrated enhancement of memory formation in human subjects by entorhinal stimulation during the learning process. The authors studied seven patients with intracranial depth electrodes who were undergoing monitoring for medically refractory epilepsy. These patients had magnetic resonance image- (MRI) guided stereotactic implantation of entorhinal or hippocampal depth electrodes, or in 4 of the 7 cases, ipsilateral implantation of both entorhinal and hippocampal electrodes. In these 4 cases hippocampal electroencephalography (EEG) data was collected during entorhinal stimulation. Patients were stimulated with parameters similar to those from other neuromodulation therapies and at levels below the threshold for inducing measureable after-discharges. Subjects were also unaware of the stimulation. They each completed a computerized spatial learning task in a virtual environment in which they had to learn the location of a number of stores and then drop off passengers at each store. Stimulation was applied to the depth electrodes as a subject was learning to navigate to particular stores (3 randomly selected stores out of 6 stores that were encountered in randomized order in each trial). Subjects completed 3 trials each in which stimulation was provided during navigation to the same 3 out of 6 stores. During a 4th trial, memory retention was tested. Spatial memory was quantified by calculating the excess path length, or the difference between the shortest possible route to a store and the actual route taken by a subject to that store. Latency, or the time elapsed between passenger pickup and dropoff at an assigned store was also measured. In each of 6 subjects, stimulation of the entorhinal cortex during the initial 3 trials resulted in a reduction of excess path length and latency for the corresponding navigation tasks in the 4th trial as compared to navigation tasks in which there had been no stimulation. The average reduction of excess path length was 64%. This was true despite the wide disparity in executive functioning pre-existing between the various patients. This effect was not observed with stimulation of hippocampal electrodes (Figure 1). In addition, for the 4 patients with both hippocampal and entorhinal electrodes, entorhinal stimulation induced greater theta-phase resetting as monitored in hippocampal electrodes as compared to non-stimulation. The large amplitude/low frequency hippocampal theta rhythm is thought to be a marker of active memory circuitry. These authors have demonstrated that entorhinal stimulation during learning can significantly enhance spatial memory, as demonstrated by improvements in this virtual world scenario performance. Although the study was limited to a small number of epilepsy patients these preliminary results are very exciting. It remains to be seen whether this kind of stimulation can be used to improve memory in patients who suffer from Alzheimer's disease or other forms of dementia and degenerative neurological conditions. Further study is needed and may reveal that stimulation in different hemispheres may result in differential improvement in verbal and spatial memory.Figure 1: Reduction in path length for the 2 different stimulated regions: A) entorhinal region, B) hippocampus. A 100% reduction would represent an excess path length of zero. Entorhinal stimulation showed consistent path length reduction, while hippocampal did not. From The New England Journal of Medicine, Suthana N, Haneef Z, Stern J, et al, Deep Brain Stimulation of Entorhinal Cortex Shows Early Promise for Enhancement of Memory Function, Volume No. 366, Pages 502-510. Copyright © (2012) Massachusetts Medical Society. Reprinted with permission from Massachusetts Medical Society.

Theoretical convergence guarantees versus numerical convergence behavior of the holomorphically embedded power flow method

Cells are the fundamental building blocks of organs and tissues. Information and mass flow through cellular contacts in these structures is vital for the orchestration of organ function. Constraints imposed by packing and cell immobility limit intercellular communication, particularly as organs and organisms scale up to greater sizes. In order to transcend transport limitations, delivery systems including vascular and respiratory systems evolved to facilitate the movement of matter and information. The construction of these delivery systems has an associated cost, as vascular elements do not perform the metabolic functions of the organs they are part of. This study investigates a fundamental trade-off in vascularization in multicellular tissues: the reduction of path lengths for communication versus the cost associated with producing vasculature. Biologically realistic generative models, using multicellular templates of different dimensionalities, revealed a limited advantage to the vascularization of two-dimensional tissues. Strikingly, scale-free improvements in transport efficiency can be achieved even in the absence of global knowledge of tissue organization. A point of diminishing returns in the investment of additional vascular tissue to the increased reduction of path length in 2.5- and three-dimensional tissues was identified. Applying this theory to experimentally determined biological tissue structures, we show the possibility of a co-dependency between the method used to limit path length and the organization of cells it acts upon. These results provide insight as to why tissues are or are not vascularized in nature, the robustness of developmental generative mechanisms and the extent to which vasculature is advantageous in the support of organ function.

This paper introduces a new approach for efficient rectangular boundary surveillance guidance for unmanned aerial vehicles (UAVs) using a continuous-curvature path described by a Lamé curve. Curvature and path length characteristics of the Lamé curve paths are analyzed, and conditions are deduced for the Lamé-curve-based circumscription of rectangular boundaries. Given the maximum turn rate capability of the UAV, a comparative study highlights significant reduction in path length as compared to the widely used elliptic circumscription. In conjunction with the desired Lamé curve path, a vector field guidance method is proposed for path-following. The vector field around the path is generated using a commanded heading angle, which encapsulates a tangent vector along the path and an arcsine shaping function of the UAV position error with respect to the desired path. Asymptotically converging behavior of the UAV position error is ascertained, and simulation studies illustrate the accurate following of the desired path for arbitrary UAV initial conditions. Indoor experimental studies utilizing a quadrotor UAV platform present the practical applicability of the proposed surveillance guidance solution.

Collision detection and avoidance are crucial for safe and efficient robotic manipulation.This paper presents an affordable collision detection system for a 3Dprinted robotic arm using a Kinect camera integrated with the Robot Operating System (ROS).The proposed approach enables real-time environment perception and collision-free trajectory generation.Experimental results show a significant improvement in collision avoidance efficiency, achieving a 40% reduction in path length while maintaining real-time performance.The computational overhead introduced by collision detection is minimal, with a 12 ms delay per frame, which can be optimized further.It is demonstrating the system's ability to effectively handle dynamic environments, such as human obstacles.This work highlights the potential of low-cost 3D sensing solutions for robotics applications, making them more accessible and scalable.

PurposeThe purpose of this paper is to solve the problem that mobile robots are still based on reactive collision avoidance in unknown dynamic environments leading to a lack of interaction with obstacles and limiting the comprehensive performance of mobile robots. A dynamic window approach with multiple interaction strategies (DWA-MIS) is proposed to solve this problem.Design/methodology/approachThe algorithm firstly classifies the moving obstacle movement intention, based on which a rule function is designed to incorporate positive incentives to motivate the robot to make correct avoidance actions. Then, the evaluation mechanism is improved by considering the time cost and future information of the environment to increase the motion states. Finally, the optimal objective function is designed based on genetic algorithm to adapt to different environments with time-varying multiparameter optimization.FindingsFaced with obstacles in different states, the mobile robot can choose a suitable interaction strategy, which solves the limitations of the original DWA evaluation function and avoids the defects of reactive collision avoidance. Simulation results show that the algorithm can efficiently adapt to unknown dynamic environments, has less path length and iterations and has a high comprehensive performance.Originality/valueA DWA-MIS is proposed, which increases the interaction capability between mobile robots and obstacles by improving the evaluation function mechanism and broadens the navigation strategy of DWA at a lower computational cost. After real machine verification, the algorithm has a high comprehensive performance based on real environment and provides a new idea for local path planning methods.

The generalized Voronoi diagram (GVD) is a powerful environment representation that defines a set of paths at maximal distance from the obstacles. Many works implicitly use this property to define safe navigation strategies for a mobile robot, but, in practice, only a few explicitly extracts in real time the GVD from the robot perception. In this paper, a real-time skeleton-based visual servoing approach is proposed for safe autonomous navigation of mobile robots in unknown structured environments. The control is based on an approximation of the local GVD using the delta medial axis, a fast and robust skeletonization algorithm. The latter produces a filtered skeleton of the freespace surrounding the robot using a pruning parameter that takes into account the robot size. A unified virtual perception frame is also introduced to project the navigable freespace obtained by possibly multiple sensors and observations. After freespace skeletonization and GVD approximation, a visual servoing control is designed and studied to ensure autonomous robot navigation along the extracted GVD branches. The approach is first tested using a simulated environment. Then, an experimental validation is carried out on a real differential wheelchair allowing the system to navigate safely and autonomously without any human intervention. Static and dynamic environment scenarios are studied. The method is immediately generalizable to any differential drive wheeled robot.

Improved path planning algorithms should minimize algorithm processing time, increase path smoothness, and shorten path length, all of which will be extremely beneficial for mobile robot traversal in large workspaces. As a result, an improved multi-objective A-star (IMOA-star) algorithm for mobile robot path planning in a large workspace was designed and implemented in Python 3.8.3 in this study. In four test cases, the proposed IMOA-star is evaluated in a large workspace with dimensions of 7120 cm × 9490 cm, and its performance is compared to the traditional A-star. When compared to the traditional A-star, the results showed that IMOA-star reduced the algorithm process time by 99.98%, improved path smoothness by 45%, reduced path length by 1.58%, and reduced the number of random points by 83.45%. Finally, the IMOA-star outperforms the traditional A-star in terms of algorithm processing time, path smoothness, path length, and the number of random points. As a result, it should be considered a viable alternative to the traditional A-star for mobile robot path planning in a large workspace.

Wheeled mobile robots moving in an unknown environment are made to face many obstacles while navigating in a planned or unplanned trajectory to reach their destination. But, no information is available regarding the failure of a leader robot of a group in both unknown and uncertain environments and the subsequent course of action by the follower robots. As the leader fails, one of the follower robots within the group can be assigned as a new leader so as to accomplish the planned trajectory. The present experimental work is carried out by a team of robot comprises of a leader robot and three follower robots and if the present leader fails, a new leader is selected from the group using leader follower approach. But, the problem of localization among multi mobile robots is subjected to communication delay and packet loss. The problem of data loss is analyzed and shows that it can be modeled as a feedback system with dual mode observers. An algorithm has been developed to compensate this packet loss during communication between controller and server in a wifi-based robotics environment. Further, the simulation results prove that the algorithm developed is efficient compared to single mode observers, for both unknown and uncertain environments among multi robots.

In this paper, we study the problem of maintaining a stable mobile robot formation, steering and localizing all robots in an unknown dynamic environment consisting of multiple periodically moving objects, without the presence of a global positioning system or a robot tracking system. We propose a distributed observer such that each agent can estimate global positions of all mobile robots and that of moving landmarks in an unknown environment. By combining the proposed distributed observer with the distributed formation control and centroid tracking control law, we show that the formation shape can be maintained by utilizing its available relative measurements and the estimated relative measurements to its neighbors, and the group's centroid follows a desired trajectory. We present stability analysis of the closed‐loop system. Finally, we validate the proposed methods in a simulation result where a group of mobile robots can maintain a robust formation and maneuver in an unknown dynamic environment.

Hybrid vehicles with hover and wing-borne forward flight capabilities offer the performance advantages of fixedwing counterparts while retaining the diverse mission capabilities of traditional rotorcraft. The ability to operate in different flight modes also presents opportunities for trajectory planning strategies that best leverage the capabilities of multi-mode aircraft. In this work, experimental flight data collected from a small tailsitter UAV operating in fixed-wing mode is used to generate energy cost heuristics for the Rapidly-exploring Random Tree (RRT*) algorithm with motion primitives based on Dubins path. From this, mission-level energy and distance statistics are generated providing insight into the vehicle-level design implications of employing a wing with variable angle of incidence. Planning parameters that were tested include the maximum number of iterations (i.e. number of nodes added to the graph), the maximum step size between nodes (i.e. graph resolution), and the energy gain for the energy cost heuristics. Results showed that for the same number of iterations, optimizing for both energy and Dubins path reduced energy consumption by 16.2% and the resultant path length by 40 meters at the higher step size. There was an average of 9.8% reduction in energy and 9.4% reduction in total path length when minimizing for both Dubins path and energy cost across all test cases. As expected, these trajectories also minimized heading changes such that new randomly sampled nodes were more likely to be added along a straight path to the goal.

To ensure the safe and efficient operation of flights in terminal maneuvering areas, airports establish standard arrival and departure procedures. However, designing these procedures is typically time-consuming and challenging for quantitative path length optimization. This paper presents an efficient optimization method for designing multiple arrival and departure procedures, taking into account the configuration of the airport and the nearby environment. The objective is to minimize the total length of routes while satisfying related operational constraints, such as obstacle avoidance and route separation. Each route is modeled in three dimensions, consisting of a horizontal polygonal line and a vertical cone. Considering the actual procedures’ merging or diverging configuration characteristics, multiple arrival or departure routes are then modeled as a tree-like topology structure. A hybrid algorithm, integrating simulated annealing and an improved A* algorithm, is developed for automated problem-solving. Simulation results, based on tests at Sweden’s Arlanda Airport, exhibit advantages over existing related research in terms of both total route length and computing time. Tests at Shanghai Pudong Airport show an 8% reduction in total path length compared to actual procedures, demonstrating the proposed method’s effectiveness and potential contributions to energy savings and emissions reduction.

Path planning in complex three-dimensional environments with multiple obstacles and narrow passages remains a critical challenge for unmanned aerial vehicles (UAVs). To address this issue, this paper proposes an improved Rapidly-exploring Dense Tree (RDT) algorithm for rapid and smooth UAV path generation. Unlike conventional RRT-based planners that rely on uniform or weakly biased sampling, the proposed method explicitly enhances local exploration efficiency while preserving the global connectivity exploration capability of the sampling tree.Specifically, the concept of density in RDT is clarified as a localized and relative sampling property, rather than a global increase in sampling nodes. Based on this definition, candidate channels are identified through three-dimensional geometric gap analysis to represent potential narrow passages, which are then used to trigger adaptive local high-density sampling. In addition, a cost-based tree expansion strategy combined with heuristic guidance is introduced to reduce redundant searches and accelerate convergence toward high-quality paths. To ensure trajectory smoothness and dynamic feasibility, a post-processing stage using cubic B-spline smoothing with curvature constraints is employed, followed by multi-constraint re-optimization considering path length, curvature, climb rate, and collision risk.Simulation experiments are conducted in two representative environments, including multi-obstacle and narrow-channel scenarios, to validate the proposed approach. Comparative studies against FS-RRT and GVP-RRT demonstrate that the proposed method achieves a 17.6% reduction in average path length and a 62.4% reduction in planning time, while generating smoother trajectories and maintaining strong obstacle avoidance capability. Additional statistical evaluations under repeated trials further confirm the improved stability and robustness of the proposed RDT framework in narrow and highly constrained environments.

PurposeThe purpose of this paper is to improve the traditional bug algorithms for the navigation of mobile robots in unknown environments by considering the limitations in previous works such as generating long path, limited to static environments as well as ignoring implementation issues. With this purpose, a new bug‐type algorithm termed Distance Histogram Bug (DH‐Bug) is proposed for overcoming these limitations.Design/methodology/approachDH‐Bug redefines the traditional motion modes and switching criteria to shorten the path length. In order to extend the framework of the traditional bug algorithms to tackle the navigation problem in dynamic environments, a new mode and the related switching conditions are designed for dealing with moving obstacles. Moreover, a realization method termed the Distance Histogram (DH) method which takes many implementation issues into full account is proposed for implementing DH‐Bug on real robots.FindingsDH‐Bug is convergent in static environments and it also guarantees “approximate” convergence based on several reasonable assumptions when there are moving obstacles. Simulations results show that DH‐Bug generates shorter average path length than some classical Bug algorithms in static environments and it also performs well in most simulations that contain moving obstacles except for some extremely adverse scenarios which have been discussed in the paper. Experiments on real robots further verify its applicability in both static environments and dynamic environments containing moving obstacles.Originality/valueCompared with previous works, DH‐Bug has three main contributions. First, in static environments, it can shorten the average path length than many classical bug algorithms in the premise of guaranteeing convergence. Second, it can be applied in dynamic environments containing moving obstacles. Third, unlike the previous bug algorithms that always ignore the practical implementation issues, DH‐Bug presents not only an abstract concept, but also a realization approach for realizing this concept on real robots.

Mobile robotics is a new field. Mobile robots range from the sophisticated space robots, to the military flying robots, to the lawn mower robots at our back yard. Mobile robotics is based on many engineering and science disciplines, from mechanical, electrical and electronics engineering to computer, cognitive and social sciences (Siegwart & Nourbakhsh, 2004). A mobile robot is an autonomous or remotely operated programmable mobile machine that is capable of moving in a specific environment. Mobile robots use sensors to perceive their environment and make decisions based on the information gained from the sensors. The autonomous nature of mobile robots is giving them an important part in our society. Mobile robots are everywhere, from military application to domestic applications. The first mobile robots as we know them today were developed during World War II by the Germans and they were the V1 and V2 flying bombs. In the 1950s W. Grey Walter developed Elmer and Elsie, two autonomous robots that were designed to explore their environment. Elmer and Elsie were able to move towards the light using light sensors, thus avoiding obstacles on their way. The evolution of mobile robots continued and in the 1970s Johns Hopkins University develops the “Beast”. The Beast used an ultrasound sensor to move around. During the same period the Stanford Cart line follower was developed by Stanford University. It was a mobile robot that was able to follow a white line, using a simple vision system. The processing was done off-board by a large mainframe. The most known mobile robot of the time was developed by the Stanford Research Institute and it was called Shakey. Shakey was the first mobile robot to be controlled by vision. It was able to recognize an object using vision, find its way to the object. Shakey, shown in Figure 1, had a camera, a rangefinder, bump sensors and a radio link. These robots had limitations due to the lack of processing power and the size of computers, and thus industrial robotics was still dominating the market and research. Industrial manipulators are attached to an off-board computer (controller) for their processing requirements and thus do not require an onboard computer for processing. Unlike industrial robots, mobile robots operate in dynamic and unknown environments and thus require many sensors (i.e. vision, sonar, laser, etc.) and therefore more processing power. Another important requirement of mobile robots is that their processing must be done onboard the moving robot and cannot be done off-board. The computer technology of the time was too bulky and too slow to meet the requirements of mobile robots. Also, sensor technology had to advance further before it could be used reliably on mobile robots.