Fig 1 - uploaded by Tomáš Krajník
Content may be subject to copyright.
Example topological map with temporal edge traversability models. Nodes are places in a map with edges linking them by movement actions. These actions can fail. i.e. door passing requires an open door, and will vary in duration depending on the environments state. The plots are illustrative, showing the predicted probability for given times p(s|t).
Source publication
In planning for deliberation or navigation in real-world robotic systems, one of the big challenges is to cope with change. It lies in the nature of planning that it has to make assumptions about the future state of the world, and the robot's chances of successively accomplishing actions in this future. Hence, a robot's plan can only be as good as...
Context in source publication
Context 1
... n.a.hawes}@cs.bham.ac.uk Based on the hypothesis that a significant amount of the changes in indoor environments are actually following certain routines, creating mostly periodic patterns of change. We propose to represent these dynamics by augmenting a topological map (see Fig. 1), composed of nodes localised in a 2D map and connecting edges with a spectral model of their traversability and action execution ...
Similar publications
This paper aims to develop a new mobile robot path planning algorithm, called generalized laser simulator (GLS), for navigating autonomously mobile robots in the presence of static and dynamic obstacles. This algorithm enables a mobile robot to identify a feasible path while finding the target and avoiding obstacles while moving in complex regions....
A numerical method based on the Pontryagin maximum principle for solving an optimal control problem with static and dynamic phase constraints for a group of objects is considered. Dynamic phase constraints are introduced to avoid collisions between objects. Phase constraints are included in the functional in the form of smooth penalty functions. Ad...
In ours paper, we have developed a Mamdani Fuzzy interference system in MATLAB R2014a for mobile robot navigation in static environment. Inputs of the fuzzy inference system are the angle between the robot orientation and the robot target orientation (i.e. robot rotation angle) and the distance measured by the ultrasonic sensor from the obstacle to...
In this paper, how to efficiently find the optimal path in complex warehouse layout and make real-time decision is a key problem. This paper proposes a new method of Proximal Policy Optimization (PPO) and Dijkstra's algorithm, Proximal policy-Dijkstra (PP-D). PP-D method realizes efficient strategy learning and real-time decision making through PPO...
We present our approach for the development, validation and deployment of a data-driven decision-making function for the automated control of a vehicle. The decisionmaking function, based on an artificial neural network is trained to steer the mobile robot SPIDER towards a predefined, static path to a target point while avoiding collisions with obs...
Citations
... Over time, it evolved from a prompt reaction to the structural changes [23] to a forecasting of changes occurring repeatedly [24]. This approach was then broadened to modeling of dynamics caused by human actions [25] and later human presence [26], people density [27], and human flows [13] in urban, human-populated environments. ...
... The first group consists of methods which attempt to model the traversability information. In contrast to methods discussed in section 6.3, these methods focus on modelling whether a given asset is reachable at the desired point in time (Haigh and Veloso 1998;Pulido Fentanes et al. 2015;Nardi and Stachniss 2020). In most cases, it is a graph-based representation where the edges model doorways or other passages that can be obstructed by obstacles. ...
Robotic mapping provides spatial information for autonomous agents. Depending on the tasks they seek to enable, the maps created range from simple 2D representations of the environment geometry to complex, multilayered semantic maps. This survey article is about maps of dynamics (MoDs), which store semantic information about typical motion patterns in a given environment. Some MoDs use trajectories as input, and some can be built from short, disconnected observations of motion. Robots can use MoDs, for example, for global motion planning, improved localization, or human motion prediction. Accounting for the increasing importance of maps of dynamics, we present a comprehensive survey that organizes the knowledge accumulated in the field and identifies promising directions for future work. Specifically, we introduce field-specific vocabulary, summarize existing work according to a novel taxonomy, and describe possible applications and open research problems. We conclude that the field is mature enough, and we expect that maps of dynamics will be increasingly used to improve robot performance in real-world use cases. At the same time, the field is still in a phase of rapid development where novel contributions could significantly impact this research area.
... These approaches assume a model is known for the human and autonomous agent. Lacerda et al. [22] plan over an MDP with transition probabilities learnt from executing each transition many times [23]. This approach does not require prior knowledge, but is limited to long-term deployment in the same environment. ...
We consider robot learning in the context of shared autonomy, where control of the system can switch between a human teleoperator and autonomous control. In this setting we address reinforcement learning, and learning from demonstration, where there is a cost associated with human time. This cost represents the human time required to teleoperate the robot, or recover the robot from failures. For each episode, the agent must choose between requesting human teleoperation, or using one of its autonomous controllers. In our approach, we learn to predict the success probability for each controller, given the initial state of an episode. This is used in a contextual multi-armed bandit algorithm to choose the controller for the episode. A controller is learnt online from demonstrations and reinforcement learning so that autonomous performance improves, and the system becomes less reliant on the teleoperator with more experience. We show that our approach to controller selection reduces the human cost to perform two simulated tasks and a single real-world task.
... They also stated that the most prominent period was most influential in a long-term experiment as it persists over a more extended time. (Fentanes et al., 2015) proposed a topological map in which there was the traversability of the edges modeled by FreMEn. They implemented the time-indexed Navigation Markov decision process that improved the planning of the navigational tasks in the changing environment. ...
... The authors that propose new mapping methods apply a wide variety of quality measuring techniques. It is not uncommon to provide only a discussion about the visual quality of the map regarding the most common directions (Kucner et al., 2013(Kucner et al., , 2016, reconstructed signals (Fentanes et al., 2015), or changes of the heat map over time (Vintr et al., 2017;Nilsang and Yuangyai, 2021). Such an approach is usually used to provide insight into the proposed concept's basic behavior. ...
... There are a lot of different opinions as to what criterion influences the acceptability of a robot in human society most (Talebpour et al., 2015;Kostavelis et al., 2017). Besides, the acceptability need not necessarily be the only point of view that defines the quality of performing the task (Krajník et al., 2015a;Fentanes et al., 2015;Kulich et al., 2016). The general idea of comparing the spatio-temporal maps is applicable and helpful in comparing different robotic models predicting diverse phenomena useful in miscellaneous robotic tasks. ...
Despite the advances in mobile robotics, the introduction of autonomous robots in human-populated environments is rather slow. One of the fundamental reasons is the acceptance of robots by people directly affected by a robot’s presence. Understanding human behavior and dynamics is essential for planning when and how robots should traverse busy environments without disrupting people’s natural motion and causing irritation. Research has exploited various techniques to build spatio-temporal representations of people’s presence and flows and compared their applicability to plan optimal paths in the future. Many comparisons of how dynamic map-building techniques show how one method compares on a dataset versus another, but without consistent datasets and high-quality comparison metrics, it is difficult to assess how these various methods compare as a whole and in specific tasks. This article proposes a methodology for creating high-quality criteria with interpretable results for comparing long-term spatio-temporal representations for human-aware path planning and human-aware navigation scheduling. Two criteria derived from the methodology are then applied to compare the representations built by the techniques found in the literature. The approaches are compared on a real-world, long-term dataset, and the conception is validated in a field experiment on a robotic platform deployed in a human-populated environment. Our results indicate that continuous spatio-temporal methods independently modeling spatial and temporal phenomena outperformed other modeling approaches. Our results provide a baseline for future work to compare a wide range of methods employed for long-term navigation and provide researchers with an understanding of how these various methods compare in various scenarios.
... A Bayesian network was demonstrated in [56], merging inputs from nonvisual sensors with object affordance characteristics obtained after object classification with CNN to calculate the certainty of the origin of the smell. Authors Kostavelis et al. [30] trained a dynamic Bayesian network for activity classification on a video on a private dataset, for human-robot interaction, attaining an average precision and recall of over 98%. ...
The simultaneous surges in the research on socially assistive robotics and that on computer vision can be seen as a result of the shifting and increasing necessities of our global population, especially towards social care with the expanding population in need of socially assistive robotics. The merging of these fields creates demand for more complex and autonomous solutions, often struggling with the lack of contextual understanding of tasks that semantic analysis can provide and hardware limitations. Solving those issues can provide more comfortable and safer environments for the individuals in most need. This work aimed to understand the current scope of science in the merging fields of computer vision and semantic analysis in lightweight models for robotic assistance. Therefore, we present a systematic review of visual semantics works concerned with assistive robotics. Furthermore, we discuss the trends and possible research gaps in those fields. We detail our research protocol, present the state of the art and future trends, and answer five pertinent research questions. Out of 459 articles, 22 works matching the defined scope were selected, rated in 8 quality criteria relevant to our search, and discussed in depth. Our results point to an emerging field of research with challenging gaps to be explored by the academic community. Data on database study collection, year of publishing, and the discussion of methods and datasets are displayed. We observe that the current methods regarding visual semantic analysis show two main trends. At first, there is an abstraction of contextual data to enable an automated understanding of tasks. We also observed a clearer formalization of model compaction metrics.
... It is relevant to also mention efforts conducted in different EU funded research projects that participate at developing safer navigation strategies for mobile robots. For instance, the STRADS project [22][23][24], is devoted to the issue of providing better predictive models of pedestrians' movements in indoor environments. The SPENCER project [25] addresses the importance of embedding in the robot's controller an understanding of social rules underlying crowd Fig. 1 Examples of robots servicing in public spaces. ...
... From pre-collision (planning actions), followed by detection, isolation, identification, classification, reaction, and post-collision response [28]. Where current state of the art has focused mostly in the first phase only (pre-collision) [22][23][24][25]27]. While detection, isolation, and identification of collisions for mobile robots is still a complex problem addressed in few works [43,44]. ...
The slogan "robots will pervade our environment" has become a reality. Drones and ground robots are used for commercial purposes while semi-autonomous driving systems are standard accessories to traditional cars. However, while our eyes have been riveted on dangers and accidents arising from drones falling and autonomous cars' crashing, much less attention has been ported to dangers arising from the imminent arrival of robots that share the floor with pedestrians and will mix with human crowds. These robots range from semi or autonomous mobile platforms designed for providing several kinds of service, such as assistant, patrolling, tour-guide, delivery, human transportation, etc. We highlight and discuss potential sources of injury emerging from contacts of robots with pedestrians through a set of case studies. We look specifically at dangers deriving from robots moving in dense crowds. In such situations, contact will not only be unavoidable but may be desirable to ensure that the robot moves with the flow. As an outlook toward the future, we also offer some thoughts on the psychological risks, beyond the physical hazards, arising from the robot's appearance and behaviour. We also advocate for new policies to regulate mobile robots traffic and enforce proper end user's training.
... Specifically, our objective is to predict where a person intends to go by simply looking at how s/he moved in the immediate past and at the shape of the environment. Our overarching goal is to use these predictions in order to create humanaware motion planning algorithm, a research area that is becoming increasingly popular [1], [2]. Our specific take is to use accurate predictions of human motions for robot plan synthesis in order to be inherently safe and compliant with unwritten social rules. ...
... Human motion predictions are generated using an embedding of the SFM into a structured neural network, i.e., a NN organised such that the neurons process the input signals according to (1). Two separate branches are designed to estimate the SFM forces in two different scenarios. ...
... wheres t k = [p t k ,v t k ] T is the state at time t, F(·) represents the second-order dynamic model (1), and A = ∂F (s t k ,f ) ∂s t k is the linearised system dynamic matrix. Moreover, Σ t k is the covariance matrix of the estimation error associated with the k-th goal state, B = ∂F (s t k ,f ) ∂f is the force linearised input vector, having additive uncertainties with covariance matrix Q (that we assume to be proportional to a perturbation in the acceleration space). ...
Generating accurate and efficient predictions for the motion of the humans present in the scene is key to the development of effective motion planning algorithms for robots moving in promiscuous areas, where wrong planning decisions could generate safety hazard or simply make the presence of the robot “socially” unacceptable. Our approach to predict human motion is based on a neural network of a peculiar kind. Contrary to conventional deep neural networks, our network embeds in its structure the popular Social Force Model, a dynamic equation describing the motion in physical terms. This choice allows us to concentrate the learning phase in the aspects which are really unknown (i.e., the model’s parameters) and to keep the structure of the network simple and manageable. As a result, we are able to obtain a good prediction accuracy even by using a small and synthetically generated training set. Importantly, the prediction accuracy remains acceptable even when the network is applied in scenarios radically different from those for which it was trained. Finally, the choices of the network are “explainable”, as they can be interpreted in physical terms. Comparative and experimental results prove the effectiveness of the proposed approach.
... The different paths that each type of vehicle can follow can be modeled by graphs. The use of graphs to model the possible paths of the vehicles has been widely used in the literature [11]. ...
In this paper, an event-Mixed Integer Linear Programming (MILP)-based algorithm is proposed to solve the task allocation problem in a Robotic Sensor Network (RSN). A fleet of two types of vehicles is considered, giving, as a result, a heterogeneous configuration of the network, since each type of vehicle has a nominal velocity and a set of allowed paths to go. The algorithm can be applied to the distributed estimation of the solar irradiance on a parabolic trough thermosolar power plant which can be used to increase the global efficiency of the plant. A simulation environment has been built to test the proposed algorithm, taking into account the behavior of the vehicles and the structure of the solar plant. The algorithm has been compared with traditional methods such as the Optimal Assignment Problem (OAP) using a set of indexes that have been defined to this purpose.
... Since the flow models are aimed to support planning unobtrusive paths, we propose to measure the quality of a given model by the number of encounters with humans, detected in the testing datasets. Similarly to [30], to reflect the ability of the models to represent the flow variations over time, we let our robots not only to plan paths but also to decide when is the best time to execute them if they have the opportunity to do so. ...
... With lower servicing ratio, the robot has more freedom to decide, when to navigate through the area and when not. That reflects the situation when the robot has to perform a certain number of tasks during the day, but it can choose the best times to perform them [30], [32]. ...
... traversability and (ii) planning paths that exploit the predictions to reduce the risk of encountering blocked passages. While existing approaches propose to make decisions according to periodic patterns of change [5], we focus on modeling spatial patterns that are independent of time information. ...
... Fremen [10] enhances a topological map with a spectral model that allows for predicting the traversability of the edges as a function of the time of day. Fentanes et al. [5] use this model for planning paths that take into account the temporal periodicity of changes in the environment. In contrast to that, we use a probabilistic approach to learn a time-agnostic model of the traversability changes on the edges of a topological representation of the environment. ...
Nowadays, mobile robots are deployed in many indoor environments such as offices or hospitals. These en- vironments are subject to changes in the traversability that often happen following patterns. In this paper, we investigate the problem of navigating in such environments over extended periods of time by capturing and exploiting these patterns to make informed decisions for navigation. Our approach uses a probabilistic graphical model to incrementally estimate a model of the traversability changes from the robot’s observations and to make predictions at currently unobserved locations. In the belief space defined by the predictions, we plan paths that trade off the risk to encounter obstacles and the information gain of visiting unknown locations. We implemented our approach and tested it in different indoor environments. The experiments suggest that, in the long run, our approach leads robots to navigate along shorter paths compared to following a greedy shortest path policy.
