Figure 1 - uploaded by Óscar Gil Viyuela
Content may be subject to copyright.
Source publication
Social robot navigation in public spaces, buildings or private houses is a difficult problem that is not well solved due to environmental constraints (buildings, static objects etc.), pedestrians and other mobile vehicles. Moreover, robots have to move in a human-aware manner—that is, robots have to navigate in such a way that people feel safe and...
Contexts in source publication
Context 1
... two techniques were tested in simulation and real-life experiments. In Social Robot navigation experiments, simulation analysis has been done in complex urban scenarios and real-life experiments were performed by the IVO robot (see Figure 1). In the IVO's robot, we use two types of sensors, a 3D LiDAR to detect obstacles, pedestrian motions and to allow the self-localization of the robot as well as an RGBD RealSense camera to detect holes and ramps. ...
Context 2
... reproduce this evaluation five times for every test environment, Figure 10. We measure the area below the curve for each of them, calculate the average proxemics score per frame through the episode and extract the mean proxemics score per frame and their standard deviation throughout the 10 episodes performed Table 4. First, as we can observe in the results, the non-linear model score in the simulated environments is clearly higher in comparison to the Linear model. ...
Context 3
... the information provided by Optitrack, the controller node can build the inputs of the model to obtain the predicted acceleration commands for the drone. The drawback of using Optitrack to accurately locate all the elements in the environment is that the working area is limited to 5 × 5 m, which imposes severe constraints on the movement of the drone and the main human, see Figure 11. First, we built a Optitrack node to detect and localize each element of the environment: the volunteer protective gear, the autonomous Drone and the obstacle of the scene; see Figure 11. ...
Context 4
... drawback of using Optitrack to accurately locate all the elements in the environment is that the working area is limited to 5 × 5 m, which imposes severe constraints on the movement of the drone and the main human, see Figure 11. First, we built a Optitrack node to detect and localize each element of the environment: the volunteer protective gear, the autonomous Drone and the obstacle of the scene; see Figure 11. With this information, the Optitrack can accurately compute the position of each required element. ...
Context 5
... this information, the Optitrack can accurately compute the position of each required element. Figure 12 provides images of the real-life experiments conducted in our laboratory. ...
Context 6
... is important to mention, however, that, due to the limitations of the experiment setup, it was not possible to fully replicate the simulated results in reality, and we had to run simplified experiments. Figure 12 shows the real-life experiments where the drone moved and was able to accompany a person. Finally, in Figure 13, we present the proxemics performance results of the real-life experiments. ...
Context 7
... 12 shows the real-life experiments where the drone moved and was able to accompany a person. Finally, in Figure 13, we present the proxemics performance results of the real-life experiments. As in the previous section, we measure the area below the curve for each experiment, and we calculate the average proxemics score per frame through different episodes performed. ...
Similar publications
To enhance collaboration between humans and robots it might be important to trigger towards humanoid robots, similar social cognitive mechanisms that are triggered towards humans, such as the adoption of the intentional stance (i.e., explaining an agents behavior with reference to mental states). This study aimed (1) to measure whether a film modul...
Citations
... In this research framework, the main focus is on developing an adaptive social force model (SFM)based navigation model, especially in the context of inclined terrain. Previously, social force navigation models have been successfully used in various applications, such as pedestrian avoidance [4]- [10], healthcare robots [11] drones [12], [13], evacuation robots [14]- [17], and navigation of soccer robots [18], [19] some also modify the SFM [20]- [22]. However, in most cases, the use of these models is limited to flat surfaces and does not consider changes that may occur to the robot during travel or navigation. ...
p> This research introduces an innovative approach to address the limitations of the commonly used social force model-based robot navigation method on flat terrain when applied to sloped terrain. The incline of the terrain becomes a crucial factor in calculating the robot’s steering output when navigating from the initial position to the target position while avoiding obstacles. Therefore, we propose a social forced model-based robot navigation system that can adapt to inclined terrain using inertial measurement unit sensor assistance. The system can detect the surface incline in real time and dynamically adjust friction and gravitational forces, ensuring the robot’s speed and heading direction are maintained. Simulation results conducted using CoppeliaSim show a significant improvement in speed adjustment efficiency. With this new navigation system, the robot can reach its destination in 59.935089 seconds, compared to the conventional social forced model which takes 63.506442 seconds, the robot is also able to reduce slip to reduce wasted movement. This method shows the potential of implementing a faster and more efficient navigation system in the context of inclined terrain. </p
... In [8], it was shown that a humanoid robot arm handing over an object with a rude or gentle attitude could influence the human interacting with the robot. While algorithms have been proposed to generate legged robot movements expressing emotions such as happy or sad [9], navigation algorithms for mobile robots focus on other dimensions such as naturalness and comfort [10][11][12][13]. These dimensions are different from social attitudes such as aggressiveness, hesitancy, or politeness, whose impact on human interactions has been studied, particularly in the field of vocal prosody [14,15]. ...
... The second part of the constraint enforces that all acceleration and deceleration phases must be followed either by a pause phase or by their opposite phase (Equation (10)), i.e., an acceleration or deceleration phase cannot be extended, since it would violate the first constraint. The increment constraint is expressed by combining these two conditions in Equation (11). ...
Many mobile robotics applications require robots to navigate around humans who may interpret the robot’s motion in terms of social attitudes and intentions. It is essential to understand which aspects of the robot’s motion are related to such perceptions so that we may design appropriate navigation algorithms. Current works in social navigation tend to strive towards a single ideal style of motion defined with respect to concepts such as comfort, naturalness, or legibility. These algorithms cannot be configured to alter trajectory features to control the social interpretations made by humans. In this work, we firstly present logistic regression models based on perception experiments linking human perceptions to a corpus of linear velocity profiles, establishing that various trajectory features impact human social perception of the robot. Secondly, we formulate a trajectory planning problem in the form of a constrained optimization, using novel constraints that can be selectively applied to shape the trajectory such that it generates the desired social perception. We demonstrate the ability of the proposed algorithm to accurately change each of the features of the generated trajectories based on the selected constraints, enabling subtle variations in the robot’s motion to be consistently applied. By controlling the trajectories to induce different social perceptions, we provide a tool to better tailor the robot’s actions to its role and deployment context to enhance acceptability.
... On the other hand, We et al. [314] proposed a pedestrian's heterogeneitybased social force model that captures the physiology and psychology attributes of pedestrians introducing physique and mentality coefficients into the SFM. Recently, SFM has also been involved in approaches integrating machine learning techniques with motion models [199,315]. ...
... The latter are rarely considered, since in research robots are usually designated for specific tasks, which influences a fragmentary approach to design and implementation. [6,9,29,40,49,54,55,59,65,73,74,80,81,92,96,101,110,111,[114][115][116]120,121,123,125,126,129,131,132,134,135,[137][138][139]141,[143][144][145][146][147][153][154][155][156][157][158][159][160]174,180,202,[204][205][206][207][208][210][211][212]220,223,227,229,[232][233][234][243][244][245][246]248,[266][267][268][269]274,276,[285][286][287]290,[298][299][300]307,308,315,317,321,323,324,[326][327][328][329][330][331][332]336,339,[341][342][343][344][345][346][348][349][350][351][352] Perceived Safety Personal spaces [9,29,49,54,59,65,73,74,80,81,101,120,123,125,129,131,132,134,137,141,[143][144][145][146][147][156][157][158][159][160]174,[205][206][207]210,212,220,223,[232][233][234][244][245][246][266][267][268][269]286,287,290,299,300,307,315,317,326,327,329,[342][343][344][345][346]348,349,352,353] O-spaces of F-formations [40,65,114,145,157,160,220,223,[232][233][234]246,[267][268][269]287,307,317,352,353] Passing speed [49,55,96,137,141,145,159,180,208,332] Motion legibility [55,74,101,139,141,147,159,160,180,202,[206][207][208]317,321,328,336,346,350] Approach direction [6,40,54,80,81,92,157,229,[244][245][246]267,269,286,307,326,332,352] Approach speed [40,54,81,92,157,245,246] Occlusion zones [132,141,266] ...
... The latter are rarely considered, since in research robots are usually designated for specific tasks, which influences a fragmentary approach to design and implementation. [6,9,29,40,49,54,55,59,65,73,74,80,81,92,96,101,110,111,[114][115][116]120,121,123,125,126,129,131,132,134,135,[137][138][139]141,[143][144][145][146][147][153][154][155][156][157][158][159][160]174,180,202,[204][205][206][207][208][210][211][212]220,223,227,229,[232][233][234][243][244][245][246]248,[266][267][268][269]274,276,[285][286][287]290,[298][299][300]307,308,315,317,321,323,324,[326][327][328][329][330][331][332]336,339,[341][342][343][344][345][346][348][349][350][351][352] Perceived Safety Personal spaces [9,29,49,54,59,65,73,74,80,81,101,120,123,125,129,131,132,134,137,141,[143][144][145][146][147][156][157][158][159][160]174,[205][206][207]210,212,220,223,[232][233][234][244][245][246][266][267][268][269]286,287,290,299,300,307,315,317,326,327,329,[342][343][344][345][346]348,349,352,353] O-spaces of F-formations [40,65,114,145,157,160,220,223,[232][233][234]246,[267][268][269]287,307,317,352,353] Passing speed [49,55,96,137,141,145,159,180,208,332] Motion legibility [55,74,101,139,141,147,159,160,180,202,[206][207][208]317,321,328,336,346,350] Approach direction [6,40,54,80,81,92,157,229,[244][245][246]267,269,286,307,326,332,352] Approach speed [40,54,81,92,157,245,246] Occlusion zones [132,141,266] ...
Navigation lies at the core of social robotics, enabling robots to navigate and interact seamlessly in human environments. The primary focus of human-aware robot navigation is minimizing discomfort among surrounding humans. Our review explores user studies, examining factors that cause human discomfort, to perform the grounding of social robot navigation requirements and to form a taxonomy of elementary necessities that should be implemented by comprehensive algorithms. This survey also discusses human-aware navigation from an algorithmic perspective, reviewing the perception and motion planning methods integral to social navigation. Additionally, the review investigates different types of studies and tools facilitating the evaluation of social robot navigation approaches, namely datasets, simulators, and benchmarks. Our survey also identifies the main challenges of human-aware navigation, highlighting the essential future work perspectives. This work stands out from other review papers, as it not only investigates the variety of methods for implementing human awareness in robot control systems but also classifies the approaches according to the grounded requirements regarded in their objectives.
... Reinforcement learning algorithms have been used to train agents that can learn optimal navigation policies. Gil et al. (2021) proposed a path planning method based on a reactive and predictive approach using reinforcement learning for robotic wheelchairs in dynamic environments. The approach combines a reactive strategy for immediate obstacle avoidance and a predictive strategy using reinforcement learning to anticipate and plan for future obstacles, enabling efficient and proactive path planning. ...
This review research paper provides a comprehensive analysis of the advancements, challenges, and methodologies in autonomous navigation and obstacle avoidance for smart robotic wheelchairs. The integration of robotics and assistive technology has revolutionized mobility solutions for individuals with impairments, enabling them to navigate complex environments independently. The paper examines the various sensor modalities, machine learning algorithms, and computer vision techniques employed for environment perception and obstacle recognition. It discusses path planning algorithms, motion control strategies, and decision-making processes for autonomous navigation. The review also addresses limitations, such as localization accuracy and dynamic environment modelling, while highlighting recent research advancements and suggesting future directions. Overall, this paper serves as a valuable resource for researchers and practitioners in the field of smart robotic wheelchairs, aiming to enhance mobility and quality of life for individuals with mobility impairments.
... [1], [2] and [7] use value networks and deep RL for learning social norms in collision avoidance. [32] compares reward functions, [6] evaluates a controller training approach, and [14] employes optimized parameters (OP-DDPG) for collision avoidance training. [9] evaluates Probabilistic roadmap (PRM-RL) on different maps. ...
... In real-world experiments, various navigation methods were evaluated. SA-CADRL [1] and GA3C-CADRL [7] policies were implemented on robotic vehicles in pedestrian environments, while [14] performed real indoor experiments with the IVO robot. [9] used a PRM-based approach, [22] employed deep RL with 3D lidar and stereo camera, and [33] tested their algorithm using a laser ranging sensor. ...
... These methods ensure human-aware navigation and achieve socially compliant in dynamic pedestrian environments. They include SA-CADRL [1],long short-term memory (LSTM) extension [7], the SOADRL algorithm [22], the neural network-based approach [21], the danger-zone prediction method proposed by [32], the approach proposed by [14] that combines machine learning techniques with the Social Force Model, the method proposed by [11] that models both human-robot cooperation and inter-human interactions, the attention mechanism-based system proposed by [28], and the approach proposed by [29] that adapts to pedestrians in real-time while ensuring trajectory planning. ...
In the past few years, there has been an increase in commercial and research focus on service robots operating in daily surroundings. These machines are anticipated to function independently in busy settings, enhancing movement efficiency and safety parameters, as well as social acceptance. Expanding conventional path planning modules to include socially aware criteria, while sustaining speedy algorithms that can adapt to human behavior without causing distress, presents a significant challenge. To address this challenge, learning methods have gained significant relevance. Among the various techniques, deep reinforcement learning, end-to-end, and inverse reinforcement learning have been the most promising. However, it is difficult to determine which techniques are superior, and sometimes, developers may obtain poor results due to inadequate data or experimental procedures during the learning stage. Therefore, it is essential to evaluate and discuss the best practices and options for an effective training stage that can improve results. As we are specifically referring to social robots, the evaluation of results should take into consideration social comfort as a key factor.
... This will be not an easy task to address for a real robot deployed in a real scenario [88]. Gil et al. [89] proposed computing robot actions by a combination of robot velocities learned by a RL model (AutoRL [90]) and robot velocities computed using an SFM [36]. ...
... The Danger Zones are formulated by considering the real time human behavior Hu et al. [87] Deep reinforcement learning framework (DRL) and the value network The DRL framework incorporating these social stress indexes Dugas et al. [88] Reinforcement Learning of Robot Navigation in Dynamic Human Environments NavRepSim environment is designed with RL applications in mind Gil et al. [89] Social Force Model (SFM) allowing human-aware Two Machine Learning techniques: Social navigation and Neural Network (NN) RL technique Francis et al. [90] PRM-RL:Probabilistic road-maps (PRMs) as the sampling-based planner and reinforcement learning-RL method in the indoor navigation context Chen et el. [80] Crowd-Robot Interaction (CRI) Attention-based Deep Reinforcement Learning Liu et al. [81] Imitation learning and deep reinforcement learning approach for motion planning in such crowded and cluttered environments Chen et al. [91] Graph convolutional network (GCN) for reinforcement learning to integrate information Attention network trained using human gaze data for assigning adjacency values. ...
In recent years, commercial and research interest in service robots working in everyday environments has grown. These devices are expected to move autonomously in crowded environments, maximizing not only movement efficiency and safety parameters, but also social acceptability. Extending traditional path planning modules with socially aware criteria, while maintaining fast algorithms capable of reacting to human behavior without causing discomfort, can be a complex challenge. Solving this challenge has involved the development of proactive systems that take into account cooperation (and not only interaction) with the people around them, the determined incorporation of approaches based on Deep Learning, or the recent fusion with skills coming from the field of human–robot interaction (speech, touch). This review analyzes approaches to socially aware navigation and classifies them according to the strategies followed by the robot to manage interaction (or cooperation) with humans.
... Autonomous navigation of social robots in different spaces is a problem that is not well solved due to the multiple drawbacks of dynamic environments, such as obstacles, pedestrians, and other mobile vehicles [15], which cause the robot to perform poorly, causing discomfort among users. However, a simple alternative solution to many applications that have this problem can be remote operation. ...
This paper presents a validation methodology for a remote system with its objective focused on a social robot. The research process starts with the customization of an application for smartphones, achieving a simple method of connection and attachment to the robot. This customization allows remote operation of the robot’s movements and an additional level of autonomy for the displacements in previously known locations. One of several teleoperations methods is the direct teleoperations method, which is used in master–slave control mode via a wireless network. Next, the article focuses on proposing a validation methodology for social robot applications design. Under this approach, two tests are performed to validate the designed application. The first one seeks to find the response speed of the communication between the robot and the mobile device wherein 10 devices with different characteristics and capabilities are used. This test is critical since a delay outside the allowable range invalidates the use of the application. The second test measures the application’s usability through a user survey, which allows for determining the preferences that people may have when using this type of application. This second test is essential to consider the overall acceptability of the social robot.
... In this paper, we present the ASP, a general planning methodology that can be used to perform different Human-Robot Collaborative Navigation (HRCN). It is flexible and can be adapted to various situations (navigation [1], accompaniment [2][3][4], approaching [5], accompaniment and approaching at the same time [6,15], or other behaviors [16]) and different robots (humanoids or drones [17,18]). In addition, the method can be improved in the future by including mechanisms that will allow specific users to customize the ASP forces and costs to include their preferences. ...
... Furthermore, other robot behaviors previously implemented had the ESFM as their core, which is part of the ASP. These methods combine the ESFM with learning algorithms or use the ESFM to achieve human-drone interaction [16][17][18]. Then, these robot behaviors can use the ASP method to include more functionalities. ...
... Furthermore, there are other methods that only include the part of the ESFM of the ASP [16][17][18], where the ESFM customization allows a Humanoid robot to perform navigation tasks that combine the ESFM with learning or enable the interactions of people with other types of robots, such as a drone. Furthermore, these methods can be improved by including the ASP or at least part of its characteristics. ...
Robots must develop the ability to socially navigate in uncontrolled urban environments to be able to be included in our daily lives. This paper presents a new robot navigation framework called the adaptive social planner (ASP) and a robotic system, which includes the ASP. Our results and previous work show that the ASP can adapt to different collaborative tasks involving humans and robots, such as independent robot navigation, human-robot accompaniment, a robot approaching people, robot navigation tasks that combine learning techniques, and human-drone interactions. Our approach in this paper focuses on demonstrating how the ASP can be customized to implement two new methods for group accompaniment: the adaptive social planner using a V-formation model to accompany groups of people (ASP-VG) and the adaptive social planner using a side-by-side model to accompany groups of people (ASP-SG). These two methods result in a robot accompanying groups of people by anticipating human and uncontrolled urban environment behaviors. Also, we develop four new robot skills to deal with unexpected human behaviors, such as rearrangement of the position of the companions inside the group, unforeseen changes in the velocity of the robot companions, occlusions among group members, and changes in the direction toward destinations in the environment. Moreover, we develop different performance metrics, based on social distances, to evaluate the tasks of the robot. In addition, we present the guidelines followed in performing the real-life experiments with volunteers, including a human-robot speech interaction to help humans create a relationship with the robot to be genuinely involved in the mutual accompaniment. Finally, we include an exhaustive validation of the methods by evaluating the behavior of the robot through synthetic and real-life experiments. We incorporate five user studies to evaluate aspects related to social acceptability and preferences of people regarding both types of robot group accompaniment.
... The Social Force Model, first proposed by Helbing in 1995 and continuously developed in the future, is a pedestrian dynamic model recognized by most scholars [7][8][9][10][11][12][13][14]. Helbing [15][16] believes that "social force" is composed of three forces: self-driving force, others' force and boundary force, whose expression is as follows: ...
Continuous pedestrian flow micro model is an important method to study pedestrian flow at present. Compared with macro model and discrete model, continuous pedestrian flow micro model can better simulate pedestrian flow phenomenon. This paper summarizes the research significance and achievements of the continuous pedestrian flow micro model. The research contents and corresponding modeling methods of relevant models are mainly introduced, and the future development of pedestrian flow micro models is prospected.
... Besides that, several applications of SFM are used for other purposes, such as companion robots (Ferrer, et al., 2013), the human leader following robot (Kuderer & Burgard, 2014), human guide robot (Dewantara & Miura, 2017) (Muallimi, et al., 2020), human-robot collision avoidance (Ratsamee, et al., 2013), social robot navigation (Kivrak, et al., 2021). Other researchers have developed SFM applications in tour-guide robots (Bellarbi, et al., 2017), healthcare robot navigation (Rifqi, et al., 2021), and SFM in quadcopter robots (Gil, et al., 2021). Another research developed SFM for robosoccer navigation purpose (Dewantara & Ariyadi, 2021) using Fuzzy-Social Force Model implemented into a ball-playing soccer robot to move from its goalpost to the opponent's goalpost without colliding with another robot. ...
ABSTRAK
Membangun sebuah sistem navigasi pada mobile robot yang bergerak di ruang sosial perlu memperhatikan beberapa aspek krusial, seperti menghindari rintangan, menjaga arah hadap robot ke tujuan, dan mencapai tujuan dengan cepat. Penelitian ini bertujuan untuk mengembangkan sistem navigasi pada Omnidirectional mobile robot menggunakan Fuzzy-Social Force Model (FSFM). Social Force Model (SFM) mampu menggerakan robot ke tujuan sambil menghindari rintangan. Fuzzy Inference System (FIS) digunakan untuk menghasilkan gain adaptif sebagai salah satu parameter SFM agar respon SFM sesuai dengan masukan dari sensor lidar. Aturan FIS dioptimasi agar mendapatkan nilai optimal menggunakan Particle Swarm Optimization (PSO). Dari hasil percobaan, mobile robot mencapai tujuan lebih cepat dengan selisih 1.59 s dan nilai error heading robot lebih kecil 0.9261 dibandingkan FSFM tanpa optimasi.
Kata kunci: Sistem Navigasi, Mobile Robot, Fuzzy-Social Force Model, Optimasi, Particle Swarm Optimization
ABSTRACT
Building a navigation system on a mobile robot moves in social space needs to consider several crucial aspects, such as avoiding obstacles, keeping the robot facing the destination, and reaching the destination quickly. This study aims to develop a navigation system on an Omnidirectional mobile robot using the Fuzzy-Social Force Model (FSFM). The Social Force Model (SFM) guides the mobile robot to its destination while avoiding obstacles. The Fuzzy Inference System (FIS) produces adaptive gain as one of the SFM parameters so that the response of the SFM matches the data of the lidar sensor. The rule base of FIS is optimized to get the optimal value using Particle Swarm Optimization (PSO). From the experimental results, mobile robots reach the destination faster with a difference of 1.59 s and a minor error in robot heading of 0.9261 compared to FSFM without optimization.
Keywords: Navigation System, Mobile Robot, Fuzzy-Social Force Model, Optimization, Particle Swarm Optimization
