No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Multilayer cognitive architecture for UAV control
Abstract
Extensive use of unmanned aerial vehicles (UAVs) in recent years has induced the rapid growth of research areas related to UAV production. Among these, the design of control systems capable of automating a wide range of UAV activities is one of the most actively explored and evolving. Currently, researchers and developers are interested in designing control systems that can be referred to as intelligent, e.g. the systems which are suited to solve such tasks as planning, goal prioritization, coalition formation etc. and thus guarantee high levels of UAV autonomy. One of the principal problems in intelligent control system design is tying together various methods and models traditionally used in robotics and aimed at solving such tasks as dynamics modelling, control signal generation, location and mapping, path planning etc. with the methods of behaviour modelling and planning which are thoroughly studied in cognitive science. Our work is aimed at solving this problem. We propose layered architecture — STRL (strategic, tactical, reactive, layered) — of the control system that automates the behaviour generation using a cognitive approach while taking into account complex dynamics and kinematics of the control object (UAV). We use a special type of knowledge representation — sign world model — that is based on the psychological activity theory to describe individual behaviour planning and coalition formation processes. We also propose path planning methodology which serves as the mediator between the high-level cognitive activities and the reactive control signals generation. To generate these signals we use a state-dependent Riccati equation and specific method for solving it. We believe that utilization of the proposed architecture will broaden the spectrum of tasks which can be solved by the UAV’s coalition automatically, as well as raise the autonomy level of each individual member of that coalition.
... Regarding the tendency toward compromise, some pieces of the literature have considered Vygotsky and Piaget to be equivalent (see e.g., Stojanov, 2001;Maia et al., 2015), neglecting their radical contradictions (see Vygotsky, 1986, p. 96). Additionally, in contrast to Vygotsky, other researchers have maintained that concepts, meaning formation, and language acquisition are, for Vygotsky, based on the direct associations among the components of experience (e.g., Billard et al., 1998;Billard and Dautenhahn, 1999;Mirolli and Parisi, 2011;Emel'yanov et al., 2016). In some models, the role of the Vygotskian socio-historical context in mental development has usually been reduced to direct external interaction among social actors (see e.g., Lindblom and Ziemke, 2003). ...
... Cultural-historical activity theory (CHAT) is not new in the context of AI. However, no studies have presented the crucial role of contradiction (see e.g., Lindblom and Ziemke, 2003;Kofod-Petersen and Cassens, 2006;O'Leary, 2008;Mirolli and Parisi, 2011;Huang and Mutlu, 2012;Suchan and Bhatt, 2012;Dhuieb et al., 2015;Maia et al., 2015;Emel'yanov et al., 2016;Gonçalves et al., 2017;Tramonte et al., 2019). It is paradoxical to instrumentally accept CHAT "without serious reflection on the complex formation process of its theoretical background" (Dafermos, 2014, p. 148). ...
By following the arguments developed by Vygotsky and employing the cultural-historical activity theory (CHAT) in addition to dialectical logic, this paper attempts to investigate the interaction between psychology and artificial intelligence (AI) to confront the epistemological and methodological challenges encountered in AI research. The paper proposes that AI is facing an epistemological and methodological crisis inherited from psychology based on dualist ontology. The roots of this crisis lie in the duality between rationalism and objectivism or in the mind-body rupture that has governed the production of scientific thought and the proliferation of approaches. In addition, by highlighting the sociohistorical conditions of AI, this paper investigates the historical characteristics of the shift of the crisis from psychology to AI. Additionally, we examine the epistemological and methodological roots of the main challenges encountered in AI research by noting that empiricism is the dominant tendency in the field. Empiricism gives rise to methodological and practical challenges, including challenges related to the emergence of meaning, abstraction, generalization, the emergence of symbols, concept formation, functional reflection of reality, and the emergence of higher psychological functions. Furthermore, through discussing attempts to formalize dialectical logic, the paper, based on contradiction formation, proposes a qualitative epistemological, methodological, and formal alternative by using a preliminary algorithmic model that grasps the formation of meaning as an essential ability for the qualitative reflection of reality and the emergence of other mental functions.
... The employment of multi-UV system in tasks, set in unknown environments, makes UV autonomous control not easy, or even impossible in some cases, to achieve. Consequently, the main trends in literature propose semiautonomous approaches, that deal with UAV flight control (Emel'yanov et al. 2016) and cooperation (Lee et al. 2013). Althnian and Agah (2016) proposed a multi-agent model for multi-robot systems to support communication strategy planning to lead agents to reach an assigned goal. ...
... A robust UV control system, that has to deal with uncertain conditions, requires the acquisition of high-level knowledge and expertise on the domain (Lee et al. 2013). Emel'yanov et al. (2016) presented a multi-level cognitive model to automate behavior generation by fusing UV articulated dynamics and kinematics. When dealing with a team of multiple UVs, the behavior generation becomes a further complicated problem to address due to the fleet management (Messous et al. 2017). ...
In recent years, the role of unmanned vehicles (UVs) is increased in many surveillance applications; they are substituting the humans in many risky activities, especially when cooperative tasks from UV team are required. To this purpose, this paper presents an agent-based framework that models a multi-UV system for surveillance applications. The agents act as wrappers for the different types of UVs, that capture data from the scene (in the area of the UV mission) and then process them, each one according to its own skills and features. The collected and processed data are then shared from the agent team to find a common agreement on the comprehension and criticality assessment of the scenario. The agent paradigm provides a seamless framework for UV interaction, making the different methodologies and technologies, designed for the different UV types, transparent. The proposal shows the agent-based modeling for a multi-UV system, where each agent hides the facilities and features of the UV it wrapped, with the aim of deploying a homogeneous interface to facilitate the collective scenario assessment in terms of critical or alerting issues, detected in the evolving scene.
... Several contributions have been done in term of control architectures [10] [13] [14] [15] [16]. beyond the known control architectures, we found the deliberative approach that has been implemented based on the sense-plan-act paradigm [17] [18]. ...
... Various control architectures have been designed in order to develop high performance systems [10], [13], [14], [15], [16]. Each of them offers new concepts in attempt to build an autonomous robot. ...
Since civil Unmanned Aerial Vehicles (UAVs) are expected to perform a wide rang of mission, the subject of designing an efficient control architecture for autonomous UAV is a very challenging problem. Several contributions had been done in order to implement an autonomous UAV. The key challenge of all these contributions is to develop the global strategy. Robotic control approaches could be classified into six categories: Deliberative, Reactive, Hybrid, Behavior, Hybrid Behavior and subsumption approach. In this paper, we will review the existing control architectures to extract the main features of civil UAVs. The definition, advantage and drawback of each architecture will be highlighted to finally provide a comparative study of the mentioned control approaches.
... A distinctive feature of the approach proposed in [1,2] is an analytical representation of a control law that eliminates online performing of expensive operations in real time: there is no needs to solve matrix SDRE for each value of state vector. The proposed results were used to construct a multi-level architecture of the unmanned aerial vehicle system [11]. However, for practical application it is important not only to ensure the stability of the original system by means of a control, but also to know the suboptimality of the proposed optimal control with respect to a particular quality criterion. ...
... The theorem 2 enables us not only to establish the fact of the uniform asymptotic stability of the closed-loop system, but also determines the criteria by which the proposed regulator is optimal. Thus, if the conditions of theorem 1 are satisfied, the estimation of suboptimality of regulator (3)-(6) is obtained from (7), (11) and if the cost function for the problem (1) ...
We investigate the nonlinear stabilizing regulator built for quasi-linear systems with coefficients depending on the state and the small parameter. The suboptimality estimation of proposed control is established using the Bellman and Lyapunov functions apparatus. This regulator is optimal close-loop control in some stabilization problem, which quality criterion is constructed in the paper.
... Behavior planning is one of the basic cognitive functions and it's effective realization is essential for all cognitive architectures used for robot control [2,5,25]. Typically planning algorithms for control systems used in robotics are developed and studied within a specific direction of Artificial Intelligence called automated (or intelligent) planning. ...
Behavior planning is known to be one of the basic cognitive functions, which is essential for any cognitive architecture of any control system used in robotics. At the same time most of the widespread planning algorithms employed in those systems are developed using only approaches and models of Artificial Intelligence and don't take into account numerous results of cognitive experiments. As a result, there is a strong need for novel methods of behavior planning suitable for modern cognitive architectures aimed at robot control. One such method is presented in this work and is studied within a special class of navigation task called smart relocation task. The method is based on the hierarchical two-level model of abstraction and knowledge representation, e.g. symbolic and subsymbolic. On the symbolic level sign world model is used for knowledge representation and hierarchical planning algorithm, PMA, is utilized for planning. On the subsymbolic level the task of path planning is considered and solved as a graph search problem. Interaction between both planners is examined and inter-level interfaces and feedback loops are described. Preliminary experimental results are presented.
... These systems are the collections of software modules automating robot behavior and are conventionally organized in a hierarchical fashion. Usually three levels of control -strategic, tactical and reactive (or named in another way but still bearing the same sense) -are distinguished [3]. In this work, we address the planning problem and examine planning methods on both strategic and tactical levels and their interaction. ...
In this paper we outline the approach of solving special type of navigation tasks for robotic systems, when a coalition of robots (agents) acts in the 2D environment, which can be modified by the actions, and share the same goal location. The latter is originally unreachable for some members of the coalition, but the common task still can be accomplished as the agents can assist each other (e.g. by modifying the environment). We call such tasks smart relocation tasks (as the can not be solved by pure path planning methods) and study spatial and behavior interaction of robots while solving them. We use cognitive approach and introduce semiotic knowledge representation - sign world model which underlines behavioral planning methodology. Planning is viewed as a recursive search process in the hierarchical state-space induced by sings with path planning signs reside on the lowest level. Reaching this level triggers path planning which is accomplished by state of the art grid-based planners focused on producing smooth paths (e.g. LIAN) and thus indirectly guarantying feasibility of that paths against agent's dynamic constraints.
... To measure the effectiveness of an approach involving embedded systems, it is necessary to have a mathematical model to validate the distributed architecture. To the best of the authors' knowledge, there is no current proposed architecture to develop UAV cognitive systems that incorporate fog concepts in itself as part of the processing stack [30,33,34] neither a specific mathematical model to validate all necessary requirements in this context [8][9][10][11][12][13][14][15]. ...
Unmanned aerial vehicles (UAVs) are a relatively new technology. Their application can often involve complex and unseen problems. For instance, they can work in a cooperative-based environment under the supervision of a ground station to speed up critical decision-making processes. However, the amount of information exchanged among the aircraft and ground station is limited by high distances, low bandwidth size, restricted processing capability, and energy constraints. These drawbacks restrain large-scale operations such as large area inspections. New distributed state-of-the-art processing architectures, such as fog computing, can improve latency, scalability, and efficiency to meet time constraints via data acquisition, processing, and storage at different levels. Under these amendments, this research work proposes a mathematical model to analyze distribution-based UAVs topologies and a fog-cloud computing framework for large-scale mission and search operations. The tests have successfully predicted latency and other operational constraints, allowing the analysis of fog-computing advantages over traditional cloud-computing architectures.
... Several works from the literature proposed architectures with cognition aspects to enhance the engagement between robotic systems and their environment [12], [15], [30]- [32]. VOLUME 11, 2023 The Intelligent Vehicle Control Architecture (IVCA) introduced in [33] facilitates cooperation among various aerial vehicles. While IVCA is designed to be adaptable, its primary focus is on collaborative aircraft operations. ...
This work presents an advanced cognitive architecture for networked aerial robots to implement autonomous swarm systems effectively. It focuses on designing, implementing, and evaluating an architecture that enables unmanned aerial vehicles (UAVs) to coordinate and cooperate for complex tasks, with or without human intervention. Inspired by artificial intelligence, cognitive science, and robotics, the architecture integrates perception, planning, decision-making, and adaptive learning to optimize swarm behavior in dynamic environments. The architecture uses a distributed processing model based on the “edge-fog-cloud” (EFC) concept. Edge-level robots handle real-time data collection, local decision-making, and environmental perception. Fog-level vehicles manage intermediate processing and supervision of the groups, while cloud servers perform comprehensive data analysis and long-term storage, being the higher-level hierarchy of the framework. This structure allows efficient distribution of computational tasks, with critical decisions made at the robot level and complex analysis done in the fog or cloud. The implementation of ARCog-NET involves deploying a multi-agent simulation system using the Robot Operating System (ROS) and Gazebo simulator, facilitating the integration of sensors, communication protocols, and data processing algorithms. The performance evaluation demonstrates the architecture’s effectiveness in a wind farm inspection scenario, where the UAV swarm exhibits improved trajectory planning, collision avoidance, and data processing efficiency. Simulation results show that ARCog-NET reduces latency, increases data throughput, and enhances operational effectiveness, providing a robust platform for future developers to focus on applications and direct robot control methods.
... For instance, the approaches of (Sanchez-Lopez et al., 2016;Molina et al., 2019) propose a Robotic Operating System (ROS) centralized software architecture to ensure a concurrent aerial mission execution with multiple UAVs. The authors in ) present a centralized cognitive architecture that integrates different methods (including those for UAV coordination) for a multi-UAV decision-making system, and the research in (Emel'yanov et al., 2016) states a multilayered control architecture to manage UAV swarm tasks. ...
This paper presents the development and implementation of a multi-UAV system focused on coverage path planning on multiple separated areas capable of re-planning the collective mission in case of unexpected events. For this purpose, we present a distributed-centralized architecture that uses heuristic and computationally efficient methods to perform the planning/re-planning and decision-making tasks during the control of the mission execution. We performed a computational evaluation of the algorithms, comparing them with other proposals, together with experiments in simulated and real flights. The results show that the system can distribute tasks equitably among the aircraft in an efficient way, even in the middle of the flight, when facing unexpected events; and show a higher computational efficiency when compared to multiple proposals in the state of the art.
... However, a precise landing without using global positioning systems (e.g., indoors) often requires a significant human intervention to ensure a successful landing. This becomes particularly important for critical missions and in situations when a UAV carries an expensive equipment onboard since most of UAV crashes occur during a landing procedure and are caused by an unexpectedly hard landing [6]. Therefore, landing assistance systems are a common way to reduce a number of incidents. ...
This paper presents a novel approach for precise UAV landing using visual sensory data. A new type of fiducial marker called embedded ArUco (e-ArUco) was developed specially for a task of a robust marker detection for a wide range of distances. E-ArUco markers are based on original ArUco markers approach and require only ArUco detection algorithms. The applicability of developed markers was validated using UAV landing experiments in a virtual environment. Both a developed marker and a landing algorithm were implemented within the ROS framework and tested in the Gazebo simulator. According to our virtual experiments, an average landing accuracy was 2.03 cm with a standard deviation of 1.53 cm.
... The creation of an intelligent automatic UAV control system, the core of which is OCE (avionics), is possible only if the control system (as a system for assessing the UAV state) has an integrated orientation system and autopilot coefficients for specific UAV flight modes. So, options should be provided for overcoming critical flight conditions, for example, from a critical roll, which may occur as a result of a gust of wind during a maneuver from a turn [55][56][57]. ...
Monitoring of air pollution using unnamed aerial systems (UAS)s is considered. The specifics and content of the information support of the measuring system using UASs are described. The mathematical model of the local atmospheric pollution field in the form of a random field is substantiated. A model of the Wiener field is considered, which describes the Brownian motion of the pollutants movements in the air. A model of the vector random field is proposed for the case of the formation of a local pollution field, the characteristics of which are estimated in the framework of the correlation theory. As an example, the analysis of the structure of a multifunctional measuring system using UASs for remote monitoring of air pollution is considered and the results of experimental studies of monitoring air pollution by radionuclides are presented, which confirms the prospect of further use of UASs for monitoring air pollution.
... In recent years, in connection with the rapid development of the direction associated with the problems of real-time control of autonomous objects motions, it became necessary to develop new algorithms for solving point-to-point control problems. In some works, researchers rely on the selection of initial constructions for the final trajectory (for example, see the trajectory planning method in [1,2]). Then the resulting initial trajectory and the corresponding control are selected taking into account the problem constraints and, as a result, an acceptable trajectory and the corresponding admissible control solving the point-to-point problem are obtained. ...
Firstly, a perturbation based correction of an admissible open-loop control in an unperturbed problem is constructed in the form of linear feedback. The constructed control transfers the perturbed system to a neighborhood of a given point. Secondly, then the resulting feedback is refined to reach the end point. The proposed algorithm is illustrated by numerical experiments.
... The intelligent vehicle control architecture (IVCA) [26] enables collaboration among multiple air vehicles. The IVCA is generic, but it aims at studying aircraft working cooperatively. ...
Big construction enterprises, such as electrical power generation dams and mining slopes, demand continuous visual inspections. The sizes of these structures and the necessary level of detail in each mission requires a conflicting set of multi-objective goals, such as performance, quality, and safety. It is challenging for human operators, or simple autonomous path-following drones, to process all this information, and thus, it is common that a mission must be repeated several times until it succeeds. This paper deals with this problem by developing a new cognitive architecture based on a collaborative environment between the unmanned aerial vehicles (UAVs) and other agents focusing on optimizing the data gathering, information processing, and decision-making. The proposed architecture breaks the problem into independent units ranging from sensors and actuators up to high-level intelligence processes. It organizes the structures into data and information; each agent may request an individual behavior from the system. To deal with conflicting behaviors, a supervisory agent analyzes all requests and defines the final planning. This architecture enables real-time decision-making with intelligent social behavior among the agents. Thus, it is possible to process and make decisions about the best way to accomplish the mission. To present the methodology, slope inspection scenarios are shown.
... Many proposals created approaches for building intelligent systems. The Aerostack [17] is one of the most prominent solutions, which presents a modular organization for command and control of autonomous, heterogeneous teams. Moreover, this approach provides an incremental development and verification of a complex system using a mixed-reality concept. ...
Efficient algorithm integration is a key issue in aerial robotics. However, only a few integration solutions rely on a cognitive approach. Cognitive approaches break down complex problems into independent units that may deal with progressively lower-level data interfaces, all the way down to sensors and actuators. A cognitive architecture defines information flow among units to produce emergent intelligent behavior. Despite the improvements in autonomous decision-making, several key issues remain open. One of these issues is the selection, coordination, and decision-making related to the several specialized tasks required for fulfilling mission objectives. This work addresses decision-making for the cognitive unmanned-aerial-vehicle architecture coined as ARCog. The proposed architecture lays the groundwork for the development of a software platform aligned with the requirements of the state-of-the-art technology in the field. The system is designed to provide high-level decision-making. Experiments prove that ARCog works correctly in its target scenario.
... To address complex multi-step decision-making problems, researchers have attempted to find methods with higher levels of intelligence. Taking into account the complex dynamics of UAVs, Emel'Yanov et al [23] proposed a cognitive architecture control system. It can solve a broad range of tasks and can raise the degree of autonomy of the control object significantly. ...
Unmanned aerial vehicles (UAVs) have played an important role in recent high-tech local wars. Seizing air control rights with UAVs will undoubtedly be a popular topic in future military development. Autonomous air combat is complex, antagonistic and mutable, and consequently, the decision-making that depends on unmanned systems is extremely challenging with very little research having been conducted on it. An intelligent air combat learning system inspired by the learning mechanisms of the brain is proposed in this paper. In accordance with research on learning, knowledge and memory, we constructed a cognitive mechanism model of the brain. Based on this model and the inferential abilities of humans, a long short-term hierarchical multi-line learning system is established. Then, the bio-inspired architecture and the basic learning principle of the system are clarified. Taking advantage of the conclusions of studies on information theory, the relationship between the knowledge updating cycle and the system learning performance is analysed. The updating cycle length adjustment problem is transformed into an optimization problem optimization problem, and system performance improvement is guaranteed. Experiments show that the system designed in this paper can acquire confrontation abilities through self-learning without prior rules; the parallel universe mechanism can significantly improve the system's learning speed when the number of parallels is within 40, and the performance of the system improves gradually and continuously. The system can master actions similar to classical tactical manoeuvres such as the high yo-yo and the barrel-roll-attack without prior knowledge. Compared with the Bayesian inference and moving horizon optimization (BI&MHO) method, the learning system proposed in this paper is more flexible in situation assessment and in the prediction of opponents' actions. Although it cannot be deployed quickly, it has a continuous learning ability.
... By being as such, these perturbations are the contradictions that arise between the agent's needs and interests, and the realization of these needs and interests in the reality, since these needs and interests open the door for "the issue of the causation and origin of our thoughts" [3]. The perturbation-based development is the crucial idea that is left out of the literature scope that mentioned Vygotsky and CHAT [e.g., 15,16,17,18,19,20]. The previous goes with Tikhomirov's consideration that "the source and motive force of this (mental) development are internal contradictions" [21]. ...
By following the guidelines of Cultural-Historical Activity Theory and the dialectic logic, this paper proposes a contradiction-based qualitative methodology for developmental robotics, in contrast with the direct associationist and quantitative aspects of the current methodologies. The proposed methodology considers that the agent-environment interaction is indirectly and dialectically structured through the emergence of meanings as the outcome of coupling the contraries through the relationship of negation. Hence, the functional qualitative reflection will be realized by abstracting the phenomenon’s substance. Moreover, a preliminary experiment of a humanoid NAO robot is represented to reflect the main ideas of the proposed framework. It is shown how mental functions can emerge based on contradictions-based development, e.g., communication and symbol emergence, concepts formation, active perception, and words acquisition. Also, it is discussed how the agent’s personification and socialization are supported.
... Furthermore, CHAT's consideration regarding the activity structure has been used in robotics to implement a "behavioral toolkit" aiding the robot to generate humanlike behavioral patterns and interaction, focusing on gaze behavior [22]. Besides [23] implemented such structure in designing a cognitive architecture for the unmanned aerial vehicles (UAVs), including the role of abstract signs and meanings, hence, covering the complex and wide range of tasks achievement in a collaboration among several UAVs. However, their definition of meaning was a direct coupling between the experience sensorimotor content, similarly to the current widely implemented definition of meaning, which also deprives the meaning of its functional content. ...
Based on the Cultural Historical Activity Theory (CHAT), the paper presents a tasks learning robotics model. The model focuses on the loop based learning, the abstract activity structure, and contradictions based functional learning through meaning emergence. Furthermore, the model introduces the “intellectual emotion” to anchor the components of learned tasks. Moreover, by proposing a preliminary experiment, the model had been discussed regarding certain robotics abilities such as functional perception and meaning emergence, abstraction and generalization, personification and interaction, including future work.
... Purposeful activity of agents both in a real environment and in a simulation implies the execution of certain actions on the manipulation of environmental objects and on moving around the map. Next, we consider the specific application of the HierMAP algorithm for the problem of smart relocation tasks [39][40][41]. ...
This paper presents a new algorithm for hierarchical case-based behavior planning in a coalition of agents – HierMAP. The considered algorithm, in contrast to the well-known planners HEART, PANDA, and others, is intended primarily for use in multi-agent tasks. For this, the possibility of dynamically distributing agent roles with different functionalities was realized. The use of a psychologically plausible approach to the representation of the knowledge by agents using a semiotic network allows applying HierMAP in groups in which people participate as one of the actors. Thus, the algorithm allows us to represent solutions of collaborative problems, forming human-interpretable results at each planning step. Another advantage of the proposed method is the ability to save and reuse experience of planning – expansion in the field of case-based planning. Such extension makes it possible to consider information about the success/ failure of interaction with other members of the coalition. Presenting precedents as a special part of the agent’s memory (semantic network on meanings) allows to significantly reduce the planning time for a similar class of tasks. The paper deals with smart relocation tasks in the environment. A comparison is made with the main hierarchical planners widely used at present.
... Situation Awareness modeling often presents solutions exploiting Cognitive Science methodologies. Multi-level cognitive models fuse knowledge on UVS complex dynamics and kinematics to automate behavior generation [21]. In multi-UVS systems, behavior generation is further complicated by the fleet management [22]. ...
In dynamic environments, autonomous and unmanned vehicle systems (UVSs) represent a reliable solution, especially when the request of high performance is a stringent constraint for complex and risky tasks, such as searching survival points, multiple target monitoring, and tracking, etc. In these cases, cooperative activities among all the involved UVSs are strategic for the achievement of a collective goal. When UVS teams work collaboratively, they collect heterogeneous data from multiple sources and bring benefits through an enhanced situational awareness (SA). Multi-UVS scenarios are, by their nature, easy to be modeled as multi-agent systems. This paper presents an agent-based modeling, governing different types of unmanned vehicles that are sent ahead in an area of interest to gather environmental, sensing, image data in order to provide a complete multi-view scenario understanding. The agent model is instantiated in each vehicle, and depending on the vehicle features, encapsulates a semantic mental modeler, customized for the specific vehicle features. The agents collect raw data from the environment and translate them into high-level knowledge, i.e., a conceptualization of the data semantics (i.e., a set of pixels assumes the meaning of a car). The proposed agent-based modeling lays on a synergy between Semantic Web technologies and Fuzzy Cognitive Map (FCM) models, producing a high-level description of the evolving scenes, and then a comprehensive scenario situational awareness.
... To measure the effectiveness of an approach involving embedded systems, it is necessary to have a mathematical model to validate the distributed architecture. To the best of the authors' knowledge, there is no current proposed architecture to develop UAV cognitive systems that incorporate fog concepts in itself as part of the processing stack [30,33,34] neither a specific mathematical model to validate all necessary requirements in this context [8][9][10][11][12][13][14][15]. ...
Unmanned aerial vehicles (UAVs) are a relatively new technology. Their application can often involve complex and unseen problems. For instance, they can work in a cooperative-based environment under the supervision of a ground station to speed up critical decision-making processes. However, the amount of information exchanged among the aircraft and ground station is limited by high distances, low bandwidth size, restricted processing capability, and energy constraints. These drawbacks restrain large-scale operations such as large area inspections. New distributed state-of-the-art processing architectures, such as fog computing, can improve latency, scalability, and efficiency to meet time constraints via data acquisition, processing, and storage at different levels. Under these amendments, this research work proposes a mathematical model to analyze distribution-based UAVs topologies and a fog-cloud computing framework for large-scale mission and search operations. The tests have successfully predicted latency and other operational constraints, allowing the analysis of fog-computing advantages over traditional cloud-computing architectures.
... Even though reactive navigation is a basic method in robotics, it is being used also for MUAV control and obstacle avoidance, as in refs. [37][38][39]. ...
This paper presents an overview of the most recent vision-based multi-rotor micro unmanned aerial vehicles (MUAVs) intended for autonomous navigation using a stereoscopic camera. Drone operation is difficult because pilots need the expertise to fly the drones. Pilots have a limited field of view, and unfortunate situations, such as loss of line of sight or collision with objects such as wires and branches, can happen. Autonomous navigation is an even more difficult challenge than remote control navigation because the drones must make decisions on their own in real time and simultaneously build maps of their surroundings if none is available. Moreover, MUAVs are limited in terms of useful payload capability and energy consumption. Therefore, a drone must be equipped with small sensors, and it must carry low weight. In addition, a drone requires a sufficiently powerful onboard computer so that it can understand its surroundings and navigate accordingly to achieve its goal safely. A stereoscopic camera is considered a suitable sensor because of its three-dimensional (3D) capabilities. Hence, a drone can perform vision-based navigation through object recognition and self-localise inside a map if one is available; otherwise, its autonomous navigation creates a simultaneous localisation and mapping problem.
... Although the full control and autonomy are essential capabilities, UAVs need to possess the high levels of SA that make them "intelligent" in order to predict future states, by the perception of the whole scenario and the comprehension of individual evolving scenes. Cognitive sciences have found a wide application in this field and layered cognitive models automate behavior generation, taking into account the complex dynamics and kinematics of the UAV [21]. Behavior generation becomes more challenging when dealing with the control of multiple UAVs (fleet management) [22]. ...
Unmanned vehicle systems are often teleoperated, semiautonomous, and strictly dependent on human operators. In complex and dynamic environments, unmanned vehicles should be autonomous in a stricter sense, which means they should exhibit a human-like behavior to be capable of accurately perceiving the environment; understanding the situation, locating and interacting with environmental elements; and reporting solutions to humans. In order to address these desiderata, a modeling of a proactive, context-aware unmanned system is presented. Precisely, the system framework is designed for an unmanned aerial vehicle (UAV) that flies over an area, and collects data in the form of video frames, sensor values, etc. It recognizes situations, senses scene object and environment data, acquires the awareness about the evolving scenes, and, finally, takes a decision based on the perception of the overall scenario. The system design is based on two primary building blocks: 1) the semantic web technologies that provide the high-level object description in the tracked scenario, and 2) the fuzzy cognitive map model that provides the cognitive accumulation of spatial knowledge in order to discern specific situations that need a decision. Although the paper presents a UAV-based surveillance system model, its applicability is shown based on a realistic case study (viz., broken car on the highway); moreover, several possible scenario configurations have been simulated to assess the criticality level perceived by the system (UAV) in a given situation and to validate the effective response/decision in the case of critical situations.
... However, the task of allocating roles in a group of autonomous agents that solve more than one of the specific problem, is universal, i.e. able to learn in the new stating the statements and simultaneously taking into account the opportunities and training of others agents that are integral parts of the solution was much more complicated and good-enough solution still does not appear. The main efforts to solve this problem of robotic system that would allow it not only to function sport, but also to learn in it, to build conceptual plans of behavior and exchange messages with other members of the coalition [18,2,6]. One of the key the subtask in creating such cognitive resources is the selection and use of method of knowledge representation and the basis for the synthesis behavior plan. ...
The paper considers the task of the group’s collective plan intellectual agents. Robotic systems are considered as agents, possessing a manipulator and acting with objects in a determined external environment. The MultiMAP planning algorithm proposed in the article is hierarchical. It is iterative and based on the original sign representation of knowledge about objects and processes, agents knowledge about themselfs and about other members of the group. For distribution actions between agents in general plan signs “I” and “Other” (“They”) are used. In conclusion, the results of experiments in the model problem “Blocksworld” for a group of several agents are presented.
... Similarly, in [17] Emel'yanov et al. propose a layered architecture of the control system that automates the RPAS behavior generation using a cognitive approach that take into account dynamics and kinematics of the aerial vehicle. The signal strength is a crucial factor if a Cloud-based architecture has to be adopted, in order to minimize the risk of disconnection. ...
... Symbolic approach to representation of knowledge and description of processes occurring in a sign world model makes it possible to solve a number of important problems in the area of situational control Osipov (1997) and Pospelov and Osipov (1997) and control of complex technical facilities Emel 'yanov, Makarov, Panov, and Yakovlev (2016). The use of sign world model to implement strategic functions of robotics systems Emel'yanov et al. (2016) demonstrates the applicability of the used approach not only to represent knowledge, but also to solve cooperative planning and role distribution problems. ...
Behavior planning is an important function of any complex technical facility intelligent control system. Presently, a symbol paradigm of artificial intelligence offers a variety of planning algorithms, including those that use precedent information, i.e. algorithms based on acquired knowledge. A symbol grounding problem within the exiting approaches of knowledge representation does not allow effective use the developed algorithms together with learning mechanisms for the purpose of solving a wide variety of applied problems by actual intelligent agents (robotics systems). This article presents the original planning algorithm (MAP Planner), which uses a sign world model as the basis for acquisition and maintenance of knowledge for future use in behavior planning. the sign problem approach describes planning as a cognitive function actualized by the world model of a subject of activity. Apart from solving symbol grounding problems and ensuring psychological and biological plausibility, a sign planning process model allows interaction of an intelligent agent with other participants in solving a cooperative task. The article presents the description of the knowledge representation method used, a MAP planning algorithm, and a model experiment in a “block world”.
... In [4], a layered architecture is proposed to control not only low and mid-level tasks but also high-level functions such as the distribution of roles in the group of UAVs, coalition formation, and behavior planning, so the UAVs are capable of performing complex tasks * Corresponding author. E-mail: lffo@ecomp.poli.br in different scenarios. ...
In this paper, we propose a hybrid mechanism based on fuzzy logics and swarm intelligence for the coordination of multiple Unmanned Aerial Vehicles (UAVs). In a previous work we have developed a finite state machine to control the swarm of robots by pre-defining expected behaviors. The main goal of our proposal is to replace the finite state machine by a Takagi-Sugeno fuzzy mechanism that smoothly changes the expected behavior of the UAVs aiming to obtain the dynamic behavior required by real-time critical systems. We deployed three different metrics to compare our proposal to the previous one, such as the target tracking capability, the anti-collision rate and the cohesion rate aiming to catch the most important aspects desired in a group of UAVs tracking targets. We performed simulations varying the number of UAVs in the environment and the number of targets to be tracked. The results indicate that the proposed mechanism diminished the number of collisions and increased the cohesion rate.
... Behavior planning is one of the basic cognitive functions and it's effective realization is essential for all cognitive architectures used for robot control [2,5,25]. Typically planning algorithms for control systems used in robotics are developed and studied within a specific direction of Artificial Intelligence called automated (or intelligent) planning. ...
Behavior planning is known to be one of the basic cognitive functions, which is essential for any cognitive architecture of any control system used in robotics. At the same time most of the widespread planning algorithms employed in those systems are developed using only approaches and models of Artificial Intelligence and don't take into account numerous results of cognitive experiments. As a result, there is a strong need for novel methods of behavior planning suitable for modern cognitive architectures aimed at robot control. One such method is presented in this work and is studied within a special class of navigation task called smart relocation task. The method is based on the hierarchical two-level model of abstraction and knowledge representation, e.g. symbolic and subsymbolic. On the symbolic level sign world model is used for knowledge representation and hierarchical planning algorithm, PMA, is utilized for planning. On the subsymbolic level the task of path planning is considered and solved as a graph search problem. Interaction between both planners is examined and inter-level interfaces and feedback loops are described. Preliminary experimental results are presented.
... These systems are the collections of software modules automating robot behavior and are conventionally organized in a hierarchical fashion. Usually three levels of control-strategic, tactical, and reactive (or named in another way but still bearing the same sense)-are distinguished [3]. In this work, we address the planning problem and examine planning methods on both strategic and tactical levels and their interaction. ...
In this paper we outline the approach of solving special type of navigation tasks for robotic systems, when a coalition of robots (agents) acts in the 2D environment, which can be modified by the actions, and share the same goal location. The latter is originally unreachable for some members of the coalition, but the common task still can be accomplished as the agents can assist each other (e.g. by modifying the environment). We call such tasks smart relocation tasks (as the can not be solved by pure path planning methods) and study spatial and behavior interaction of robots while solving them. We use cognitive approach and introduce semiotic knowledge representation - sign world model which underlines behavioral planning methodology. Planning is viewed as a recursive search process in the hierarchical state-space induced by sings with path planning signs reside on the lowest level. Reaching this level triggers path planning which is accomplished by state of the art grid-based planners focused on producing smooth paths (e.g. LIAN) and thus indirectly guarantying feasibility of that paths against agent's dynamic constraints.
p>The paper considers the problem of predicting the agent's activity based on the text description of the task and visual analysis of the environment. An update of the approaches of classical cognitive architecture is proposed, allowing its application in a real environment. An addition to the semiotic method of symbolic designation with the author's neural network mechanism for linking vectors of text and visual spaces is developed. A number of experiments with the obtained model in a complex environment of a car driving emulator are conducted.</p
This paper presents the development and implementation of a multiple unmanned aerial vehicle system focused on coverage path planning on multiple separated areas capable of re‐planning the collective mission in case of unexpected events. For this purpose, we present a distributed‐centralized architecture that uses heuristic and computationally efficient methods to perform the planning/re‐planning and decision‐making tasks during the control of the mission execution. We performed a computational evaluation of the algorithms, comparing them with other proposals, together with experiments in simulated and real flights. The results show that the system can distribute tasks equitably among the aircraft in an efficient way, even in the middle of the flight, when facing unexpected events; and show a higher computational efficiency when compared to multiple proposals in the state of the art.
Driven by new military requirements and technological innovation, modern warfare is developing towards system-of-systems combat based on distributed operations. Correspondingly, the traditional kill chains are extending to kill webs or effects webs. During this process, unmanned aircraft systems play a more and more important role, especially those networked, small, low-cost ones, and are collaborated and integrated as goal-driven systems in the new distributed combat style. In this paper, we firstly introduce the shift from kill chain to kill web, which has put forward new requirements for the applications of UAS. Then, typical UAS distributed dynamic cooperative combat styles are discussed. Under this background, several key issues of cooperative control and management for UAS that need to be addressed by future research activities are analyzed and summarized, related with cooperation architecture, communication, navigation, decision-making, planning, guidance, and control. Finally, a conclusion is briefly summarized with future development tendency and challenges.
Unmanned aerial vehicles (UAVs) are considered to be recent developing technology having ability of making various basic public and private processes better. UAVs help in speeding up the recovery and rescue activities and also helpful in delivery systems. Development of UAV systems by traditional techniques needs large amount of effort, time, and cost. Management of UAV systems using conventional techniques has many limitations. As the UAV airspace turn out to be more restrictive, it is significant to manage UAV systems by utilizing innovative methods of collision avoidance. In this chapter, cloud-web-application study is presented which offers real-time flight managing and regulating UAVs. For every linked UAV, exhaustive reading of the accelerometer UAV sensor, GPS and provides ultrasonic sensor as well as visual cameras in addition to status reports from small inner elements of UAVs. The flexible map overlap looks at current routes and existing UAV sites, letting users easily manage all aircraft operations. The system perceives and avoids major impacts by auto-adjusting the UAV flight modes and notifying users about changes. The main aim is to streamline efforts, reduction in time, and cost of collaborative applications for UAVs which are heterogenous. This chapter deliberates how UAVs are unified with the cloud computing standard for providing universal access to their services. UAV resources are tracked in a novel flexible web-service design and has integrated communication amid services known as resource oriented architecture. The UAV-cloud framework to facilitate the design of pervasive UAV viewpoints, also the trivial integrated design of UAV resources is also discussed in this chapter.KeywordsUAVsResource oriented architectureRESTfulCloud computing
The climate and weather dynamics in the past few years has driven a massive increase in the number and intensity of flood disasters, which severely claim casualties in human, goods and properties. Aimed to reduce these casualties, emerging software-defined internet protocol-based communication technologies in the form of Internet of Things (IoT) have attracted strong interests from disaster mitigation stakeholders to rapidly locate victims and acquire their relevant information, which in turn can boost up the efficiency and effectiveness of Search and Rescue (SAR) missions. In order to capture state-of-the-art development and technological challenges, this paper presents an extensive review on the flood SAR systems, highlighting some of the key emerging IoT technologies that prove or are potentially useful in improving the SAR operation by the rescuers. Furthermore, a comprehensive study on different existing communication technologies for SAR is provided, covering the system architecture, communication network compositions and applications. Based on the critical analysis of existing works, this paper puts forward a proposal on an IoT-aided integrated flood management framework to support SAR in the flood-catchment areas, leveraging upon three-domain (ground, water and air) collaborative wireless networks.
This paper considers the cutting-edge scientific problem of controlling unmanned aerial vehicles (UAVs) in unstable conditions based on intelligent-geometric theory that combines geometric control methods (methods of optimal control, complex motion control and stabilization, trajectory tracking, differential pursuit-evasion games, etc.) with intelligent control methods using tools of artificial intelligence (productions, semantic networks, fuzzy logic, frame-based behavioral microprograms and operations, methods of knowledge acquisition, etc.). Such integration provides reliable and high-performance control techniques for operating in uncertain environments under wind disturbances. Hierarchical architecture of intelligent-geometric control system is proposed, designed for joint application of precise geometric and adaptive intelligent control methods as parts of a single robotic system. In accordance with the proposed architecture, the solution to the problem of controlling a UAV group taking into account mathematical models of an aircraft and wind loads was simulated in MATLAB Simulink system.
In the paper, we consider the problem of the robotic movement inaccuracy. We suggest that clarifying the abstract actions of the behavior planner will help build more precise control of the robot. A multi-agent planner for the synthesis of the abstract actions with refinement for a two-dimensional movement task was proposed. We analyze the problem of the action execution by robots and present a way to solve navigation problems through the use of reinforcement learning and deep learning algorithms. This method made it possible to synthesize sets of atomic sub-actions for correcting the state of the robotic platform at each moment of time. We conducted a set of experiments in single and multiagent settings. The synthesis of sub-actions of several tasks formed a training set for the RL model, which was tested on a test example. All considered tasks consisted of moving robotic platforms across a map with obstacles and manipulating environmental objects.
We present a model of Reinforcement Learning, which consists of modified neural-network architecture with spatio-temporal connections, known as Temporal Hebbian Self-Organizing Map (THSOM). A number of experiments were conducted to test the model on the maze solving problem. The algorithm demonstrates sustainable learning, building a near to optimal routes. This work describes an agents behavior in the mazes of different complexity and also influence of models parameters at the length of formed paths.
This paper introduces a Montecarlo Genetic Algorithm, hierarchical, multi-objective optimization of a Vertical Take Off landing Unmanned Aerial Vehicle having a tail sitter configuration. An optimization of the hierarchical type is introduced in place of the methods generally used multi-objective optimization, such as Pareto and “arbitrary” weighted sums. A Montecarlo method optimizes the weights of the final objective function used by the Genetic Algorithm. A very simple “spreadsheet based” algorithm defines the CAD model of the Genetic Algorithm individuals in order to evaluate the performance of the candidates. The optimization method described in this study appears to be very effective. Then experimental tests were conducted with scaled-down prototypes. Four flight tests were performed: Take Off, Cruise, Slow flight, Landing. A Taguchi matrix was defined for each experiment. The tests started from a prototype that comes directly from the Montecarlo Genetic Algorithm optimization and led to the final prototype shown along the paper (page 7, right figure). Unfortunately, the tail sitter approach proved poor control authority in the final phase of the vertical landing. Even the “final” prototype showed unsatisfactory behavior in case of erratic wind gusts. This unsolved problem is common to the tail sitter configuration that requires a power control by air jets or additional propeller to control the aircraft in the final phase of landing. Unfortunately, this necessity renders the tail sitter configuration inconvenient for small Unmanned Aerial Vehicles.
M. G. Dmitriev, D. A. Makarov
"Smooth nonlinear controller in a weakly nonlinear control system with state depended coefficients"
Work is devoted to the problem of constructing a smooth nonlinear stabilizing controller for a class of control systems, in which the coefficients are weakly nonlinear functions of the state. This weakly state dependency formalized in the presence of the so-called small parameter in the right-hand side of the system dynamics equations. The algorithm, allowing building the controller in numerical-analytical form and significantly reducing the computational costs, is proposed (in Russian).
Functions that are referred in psychology as functions of consciousness are considered. These functions include reflection, consciousness of activity motivation, goal setting, synthesis of goal oriented behavior, and some others. The description is based on the concept of sign, which is widely used in psychology and, in particular, in the cultural-historical theory by Vygotsky, in which sign is interpreted informally. In this paper, we elaborate upon the concept of sign, consider mechanisms of sign formation, and some self-organization on the set of signs. Due to the work of self-organization mechanisms, a new method for the representation of the world model of an actor appears. The concept of semiotic network is introduced that is used for the examination of the actor’s world models. Models of some functions indicated above are constructed. The second part of the paper is devoted to functions of self-consciousness and to the application of the constructed models for designing plans and constructing new architectures of intelligent agents that are able, in particular, to distribute roles in coalitions.
In this paper a possible general framework for the representation of concepts in cognitive artificial systems and cognitive architectures is proposed. The framework is inspired by the so called proxytype theory of concepts and combines it with the heterogeneity approach to concept representations, according to which concepts do not constitute a unitary phenomenon. The contribution of the paper is twofold: on one hand, it aims at providing a novel theoretical hypothesis for the debate about concepts in cognitive sciences by providing unexplored connections between different theories; on the other hand it is aimed at sketching a computational characterization of the problem of concept representation in cognitively inspired artificial systems and in cognitive architectures.
Over long lifetimes, learning agents accumulate large stores of knowledge. To support human-level decision-making, their cognitive architectures must efficiently manage this experience and bring to bear pertinent data to act in the world. 1 Prior psychological and computational work suggests the need for multiple, dissociated memory systems, citing significant functional and computational tradeoffs that arise when implementing a single memory mechanism for different types of learning tasks. In this context, we develop a memory-centric analysis of Soar 9, a general cognitive architecture that incorporates multiple long-term memories. In this analysis, we explore the functional abilities, computational opportunities, and theoretical challenges entailed by integrating a diverse set of symbolic memory systems.
Grids with blocked and unblocked cells are often used to represent terrain in
robotics and video games. However, paths formed by grid edges can be longer
than true shortest paths in the terrain since their headings are artificially
constrained. We present two new correct and complete any-angle path-planning
algorithms that avoid this shortcoming. Basic Theta* and Angle-Propagation
Theta* are both variants of A* that propagate information along grid edges
without constraining paths to grid edges. Basic Theta* is simple to understand
and implement, fast and finds short paths. However, it is not guaranteed to
find true shortest paths. Angle-Propagation Theta* achieves a better worst-case
complexity per vertex expansion than Basic Theta* by propagating angle ranges
when it expands vertices, but is more complex, not as fast and finds slightly
longer paths. We refer to Basic Theta* and Angle-Propagation Theta*
collectively as Theta*. Theta* has unique properties, which we analyze in
detail. We show experimentally that it finds shorter paths than both A* with
post-smoothed paths and Field D* (the only other version of A* we know of that
propagates information along grid edges without constraining paths to grid
edges) with a runtime comparable to that of A* on grids. Finally, we extend
Theta* to grids that contain unblocked cells with non-uniform traversal costs
and introduce variants of Theta* which provide different tradeoffs between path
length and runtime.
This book explores the idea that we have two minds – one being automatic, unconscious, and fast, the other controlled, conscious, and slow. In recent years, there has been great interest in so-called dual-process theories of reasoning and rationality. According to such theories, there are two distinct systems underlying human reasoning: an evolutionarily old system that is associative, automatic, unconscious, parallel, and fast; and a more recent, distinctively human system which is rule-based, controlled, conscious, serial, and slow. Within the former, processes are held to be innate and to use heuristics that evolved to solve specific adaptive problems. In the latter, processes are taken to be learned, flexible, and responsive to rational norms. Despite the attention these theories are attracting, there is still poor communication between dual-process theorists themselves, and the substantial bodies of work on dual processes in cognitive psychology and social psychology remain isolated from each other. The book brings together researchers on dual processes to summarize the latest research, highlight key issues, present different perspectives, explore implications, and provide a stimulus to further work. It includes new ideas about the human mind both by contemporary philosophers interested in broad theoretical questions about mental architecture, and by psychologists specializing in traditionally distinct and isolated fields.
Connectivity determines the function of neural circuits. Historically, circuit mapping has usually been viewed as a problem of microscopy, but no current method can achieve high-throughput mapping of entire circuits with single neuron precision. Here we describe a novel approach to determining connectivity. We propose BOINC ("barcoding of individual neuronal connections"), a method for converting the problem of connectivity into a form that can be read out by high-throughput DNA sequencing. The appeal of using sequencing is that its scale-sequencing billions of nucleotides per day is now routine-is a natural match to the complexity of neural circuits. An inexpensive high-throughput technique for establishing circuit connectivity at single neuron resolution could transform neuroscience research.
This paper briefly describes a robot architecture that has been under development for the last eight years. This architecture uses several levels of abstraction and description languages that are compatible between levels. The makeup of the architecture helps to coordinate planful activities with real-time behaviors for dealing with dynamic environments. In recent years, many architectures have been created with similar attributes. The two features that distinguish this architecture from most of those are: 1) a variety of useful software tools have been created to help implement this architecture on multiple real robots; and 2) this architecture, or parts of it have been implemented on over half a dozen very different robot systems using a variety of processors, operating systems, effectors and sensor suites.
This paper describes a real-time motion planning algorithm, based on the rapidly-exploring random tree (RRT) approach, applicable to autonomous vehicles operating in an urban environment. Extensions to the standard RRT are predominantly motivated by: 1) the need to generate dynamically feasible plans in real-time; 2) safety requirements; 3) the constraints dictated by the uncertain operating (urban) environment. The primary novelty is in the use of closed-loop prediction in the framework of RRT. The proposed algorithm was at the core of the planning and control software for Team MIT's entry for the 2007 DARPA Urban Challenge, where the vehicle demonstrated the ability to complete a 60 mile simulated military supply mission, while safely interacting with other autonomous and human driven vehicles.
In this paper, we examine the motivations for research on cognitive architectures and review some candidates that have been explored in the literature. After this, we consider the capabilities that a cognitive architecture should support, some properties that it should exhibit related to representation, organization, performance, and learning, and some criteria for evaluating such architectures at the systems level. In closing, we discuss some open issues that should drive future research in this important area.
Pathfinding in uniform-cost grid environments is a problem
commonly found in application areas such as robotics and
video games. The state-of-the-art is dominated by hierar-
chical pathfinding algorithms which are fast and have small
memory overheads but usually return suboptimal paths. In
this paper we present a novel search strategy, specific to grids,
which is fast, optimal and requires no memory overhead. Our
algorithm can be described as a macro operator which iden-
tifies and selectively expands only certain nodes in a grid
map which we call jump points. Intermediate nodes on a
path connecting two jump points are never expanded. We
prove that this approach always computes optimal solutions
and then undertake a thorough empirical analysis, comparing
our method with related works from the literature. We find
that searching with jump points can speed up A* by an order
of magnitude and more and report significant improvement
over the current state of the art.
This paper describes an implementation of the 3T robot architecture which has been under development for the last eight years. The architecture uses three levels of abstraction and description languages which are compatible between levels. The makeup of the architecture helps to coordinate planful activities with real-time behaviours for dealing with dynamic environments. In recent years, other architectures have been created with similar attributes but two features distinguish the 3T architecture : (1) a variety of useful software tools have been created to help implement this architecture on multiple real robots; and (2) this architecture, or parts of it, have been implemented on a variety of very different robot systems using different processors, operating systems, effecters and sensor suites.
One approach in pursuit of general intelligent agents has been to concentrate on the underlying cognitive architecture, of which Soar is a prime example. In the past, Soar has relied on a minimal number of architectural modules together with purely symbolic representations of knowledge. This paper presents the cognitive architecture approach to general intelligence and the traditional, symbolic Soar architecture. This is followed by major additions to Soar: non-symbolic representations, new learning mechanisms, and long-term memories.
The identification of performance-optimizing parameter settings is an
important part of the development and application of algorithms. We describe an
automatic framework for this algorithm configuration problem. More formally, we
provide methods for optimizing a target algorithm's performance on a given
class of problem instances by varying a set of ordinal and/or categorical
parameters. We review a family of local-search-based algorithm configuration
procedures and present novel techniques for accelerating them by adaptively
limiting the time spent for evaluating individual configurations. We describe
the results of a comprehensive experimental evaluation of our methods, based on
the configuration of prominent complete and incomplete algorithms for SAT. We
also present what is, to our knowledge, the first published work on
automatically configuring the CPLEX mixed integer programming solver. All the
algorithms we considered had default parameter settings that were manually
identified with considerable effort. Nevertheless, using our automated
algorithm configuration procedures, we achieved substantial and consistent
performance improvements.
The graph-search algorithms developed between 60s and 80s were widely used in many fields, from robotics to video games. The
A* algorithm shall be mentioned between some of the most important solutions explicitly oriented to motion-robotics, improving
the logic of graph search with heuristic principles inside the loop. Nevertheless, one of the most important drawbacks of
the A* algorithm resides in the heading constraints connected with the grid characteristics. Different solutions were developed
in the last years to cope with this problem, based on post-processing algorithms or on improvements of the graph-search algorithm
itself. A very important one is Theta* that refines the graph search allowing to obtain paths with “any” heading. In the last
two years, the Flight Mechanics Research Group of Politecnico di Torino studied and implemented different path planning algorithms.
A Matlab based planning tool was developed, collecting four separate approaches: geometric predefined trajectories, manual
waypoint definition, automatic waypoint distribution (i.e. optimizing camera payload capabilities) and a comprehensive A*-based
algorithm used to generate paths, minimizing risk of collision with orographic obstacles. The tool named PCube exploits Digital
Elevation Maps (DEMs) to assess the risk maps and it can be used to generate waypoint sequences for UAVs autopilots. In order
to improve the A*-based algorithm, the solution is extended to tri-dimensional environments implementing a more effective
graph search (based on Theta*). In this paper the application of basic Theta* to tri-dimensional path planning will be presented.
Particularly, the algorithm is applied to orographic obstacles and in urban environments, to evaluate the solution for different
kinds of obstacles. Finally, a comparison with the A* algorithm will be introduced as a metric of the algorithm performances.
Adaptive control of thought-rational (ACT-R; J. R. Anderson & C. Lebiere, 1998) has evolved into a theory that consists of multiple modules but also explains how these modules are integrated to produce coherent cognition. The perceptual-motor modules, the goal module, and the declarative memory module are presented as examples of specialized systems in ACT-R. These modules are associated with distinct cortical regions. These modules place chunks in buffers where they can be detected by a production system that responds to patterns of information in the buffers. At any point in time, a single production rule is selected to respond to the current pattern. Subsymbolic processes serve to guide the selection of rules to fire as well as the internal operations of some modules. Much of learning involves tuning of these subsymbolic processes. A number of simple and complex empirical examples are described to illustrate how these modules function singly and in concert.
This paper addresses the problem of simultaneous localization and mapping (SLAM) using vision-based sensing. We present and analyse an implementation of a Rao- Blackwellised particle filter (RBPF) that uses stereo vision to localize a camera and 3D landmarks as the camera moves through an unknown environment. Our implementation is robust, can operate in real-time, and can operate without odometric or inertial measurements. Furthermore, our approach supports a 6-degree-of-freedom pose representation, vision-based ego-motion estimation, adaptive resampling, monocular operation, and a selection of odometry-based, observation-based, and mixture (combining local and global pose estimation) proposal distributions. This paper also examines the run-time behavior of efficiently designed RBPFs, providing an extensive empirical analysis of the memory and processing characteristics of RBPFs for vision-based SLAM. Finally, we present experimental results demonstrating the accuracy and efficiency of our approach.
Pathfinding in uniform-cost grid environments is a problem commonly found in application areas such as robotics and video games. The state-of-the-art is dominated by hierarchical pathfinding algorithms which are fast and have small memory overheads but usually return suboptimal paths. In this paper we present a novel search strategy, specific to grids, which is fast, optimal and requires no memory overhead. Our algorithm can be described as a macro operator which identifies and selectively expands only certain nodes in a grid map which we call jump points. Intermediate nodes on a path connecting two jump points are never expanded. We prove that this approach always computes optimal solutions and then undertake a thorough empirical analysis, comparing our method with related works from the literature. We find that searching with jump points can speed up A* by an order of magnitude and more and report significant improvement over the current state of the art.
In the article we solve path planning task for an agent being multirotor unmanned aerial vehicle (multicopter). We propose an approach of estimating path geometry constraints based on UAV flight dynamics model and control constraints. Than we introduce a new path finding method which takes into consideration those geometry constraints and study this method both theoretically and empirically.
The hypothesis is advanced that the events of the subjective experience emerge as a result of the synthesis of three kinds of information in the cortical areas which are crucial for this mental function. The three kinds of information are presented by the sensory inputs, reproduction from memory, and that from motivational centers. The hypothesis is based on the studies of brain mechanisms of perception and thinking. It has been shown that sensation emerges as a result of the synthesis of information about the physical parameters and significance of the stimulus performed in the projection cortex neurons. This synthesis is provided by the circular run of nerve impulses from the projection cortex to associative cortex, then to hippocampus and the hypothalamic motivation centers with subsequent return to the projection cortex. It has been shown also that in thinking operations the cortical connections converge to definite centers named the interaction foci. Their topography is specific for particular thinking operations. Thus, in imaginative thinking the foci are located in the temporo-parietal areas and in abstract-verbal thinking in the frontal cortex. It is suggested that comparison and synthesis of information in the interaction foci result in decision making.
The physical, mechanical and freezing and thawing properties of non cement porous vegetation concrete using non-sintering inorganic binder have been evaluated in this study. Four types of porous vegetation concrete according to the binder type is evaluated. The pH value, void ratio, compressive strength, repeated freezing and thawing properties were tested. The test results indicate that the physical, mechanical and repeated freezing and thawing properties of porous vegetation concrete using the non-sintering inorganic binder is increased or equivalent compared to the porous vegetation concrete using the blast furnace slag + cement and hwang-toh + cement binders. Also, Vegetation monitoring test results indicate the porous vegetation concrete using the non-sintering inorganic binder have increasing effects of vegetation growth.
A nonlinear control problem has been posed by Bupp et al.1 to provide a benchmark for evaluating various nonlinear control design techniques. In this paper, the capabilities of the state-dependent Riccati equation (SDRE) technique are illustrated in producing two control designs for the benchmark problem. The SDRE technique represents a systematic way of designing nonlinear regulators. The design procedure consists of first using direct parameterization to bring the nonlinear system to a linear structure having state-dependent coefficients (SDC). A state-dependent Riccati equation is then solved at each point x along the trajectory to obtain a nonlinear feedback controller of the form u = - R-1(x)BT(x)P(x)x, where P(x) is the solution of the SDRE. Analysis of the first design shows that in the absence of disturbances and uncertainties, the SDRE nonlinear feedback solution compares very favorably to the optimal open-loop solution of the posed nonlinear regulator problem, the latter being obtained via numerical optimization. It is also shown via simulation that the closed-loop system has stability robustness against parametric variations and attenuates sinusoidal disturbances. In the second design it is demonstrated how a hard bound can be imposed on the control magnitude to avoid actuator saturation.
Square grids are commonly used in robotics and game development to model an
agent's environment, and well known in Artificial Intelligence heuristic search
algorithms (A*, JPS, Theta* etc.) are utilized for grid path planning. A lot of
research in this area has been focused so far on finding the shortest paths
while in many applications producing smooth paths is preferable. In our work,
we study the problem of generating smooth grid paths and concentrate on angle
constrained path planning. We put angle constrained path planning problem
formally and present a new algorithm of solving it - LIAN. We examine LIAN both
theoretically and empirically. On the theoretical side, we prove that LIAN is
sound and complete (under well-defined restrictions). On the experimental side,
we show that LIAN significantly outperforms competitors in ability to find
solutions under tough resource constraints and in computational efficiency.
This paper exposes the Navigation and Control technology embedded in a recently commercialized micro Unmanned Air Vehicle (UAV), the AR.Drone, which cost and performance are unprecedented among any commercial product for mass markets. The system relies on state-of-the-art indoor navigation systems combining low-cost inertial sensors, computer vision techniques, sonar, and accounting for aerodynamics models.
This study suggests a new approach based on Theta* algorithm to create paths in real-time, considering both angular rate (yaw rate) and heading angle of unmanned surface vehicles (USVs). Among the various path planning algorithms, a grid map-based path planning algorithm has frequently been employed for maintaining the control stability of nonholonomic USVs. This algorithm is powerful in that it generates a path with the fastest computation time. Most path planning algorithms for the ocean environment, however, have only been performed on an Euclidean group SE(2) grid map with no considerations of the vehicle performance. Other path planning algorithms for an SE(3) grid map generate resultant paths with consideration of vehicle heading, but most of them do not operate in real-time. Moreover, these algorithms always generate constant angle at each orientation node, and the size of graph must be expanded to provide finer angular resolution. To solve these problems, the angular rate-constrained Theta* (ARC-Theta*) is proposed to create paths in real-time with considerations of vehicle performance on the SE(2) weighted occupancy grid map. The performance of this proposed algorithm is verified with simulations and experiments. The results showed that the ARC-Theta* algorithm is effective for global path planning of USVs.
Research in robotic systems has traditionally been identified with approaches that are characterized by the use of carefully crafted representations and processes to find optimal solutions. The use of such representations and processes, which we refer to as the algorithmic approach, is uniquely suited for problems requiring strong models, i.e., tasks and domains that are well defined, and/or involve close interaction with the environment. These problems have historically been the focus of robotics research because they exercise perceptual, motor and manipulation capabilities that form the basic foundational abilities required for every robotic agent. Recent work (for example ROS and Tekkotsu) on the abstraction and encapsulation of perception and motor functionality has standardized the above mentioned foundational abilities and allowed researchers to study problems in less clearly defined and open-ended domains: problems that have previously been considered the province of AI and Cognitive Science. In this paper, we argue that the study of these problems (examples of which include multi-agent interaction, instruction following and reasoning in complex domains) referred to under the rubric of Cognitive Robotics is best achieved via the use of cognitive architectures – unified computational frameworks developed specifically for general problem solving and human cognitive modeling. We lay out the relevant architectural concepts and principles and illustrate them using nine cognitive architectures that are under active development – Soar, ACT-R, CLARION, GMU-BICA, Polyscheme, Co-JACK, ADAPT, ACT-R/E, and SS-RICS.
The computationally efficient search for robust feasible paths for unmanned aerial vehicles (UAVs) in the presence of uncertainty is a challenging and interesting area of research. In uncertain environments, a “conservative” planner may be required but then there may be no feasible solution. In this paper, we use a chance constraint to limit the probability of constraint violation and extend this framework to handle uncertain dynamic obstacles. The approach requires the satisfaction of probabilistic constraints at each time step in order to guarantee probabilistic feasibility. The rapidly-exploring random tree (RRT) algorithm, which enjoys the computational benefits of a sampling-based algorithm, is used to develop a real-time probabilistically robust path planner. It incorporates the chance constraint framework to account for uncertainty within the formulation and includes a number of heuristics to improve the algorithm’s performance. Simulation results demonstrate that the proposed algorithm can be used for efficient identification and execution of probabilistically safe paths in real-time.
Abstract UAVs offer tantalizing capabilities to the warfighter, such as tireless observation, quick recognition, and rapid reaction to today’s changing battlespace. These trends are important because they aid Warfighter in their duties. Today, unmanned systems exist that extend the vision and the reach of the Warfighter. However, they spend so much time managing these assets that they lose effectiveness as a Warfighter. This is a particular problem if the warfighter’s role is one demanding continuous sensory and mental workload, such as the Co-Pilot/Gunner (CPG) of an Apache Longbow attack helicopter. Autonomy, the ability of vehicles to conduct most of their operation without human supervision, can help relieve the burden ofproviding continuous oversight of the UAV’s operation. This moves the Warfighter’s role from control to command, enabling them to perform their duties more effectively and successfully. Collaboration, the ability of teams of vehicles to coordinate their activities without human oversight, moves unmanned systems to the level of a true force multiplier, giving a single human warfighter the power of multiple coordinated, intelligent platforms. Introduction, Lockheed Martin has developed ,a general ,architecture for Collaborative Autonomy ,that provides both the Autonomy and,the Collaboration necessary ,to achieve ,this force multiplication. This architecture provides the capability for individual unmanned ,vehicles to operate with unparalleled degrees of intelligence and autonomy, and for groups of unmanned,vehicles to operate ,together effectively as a team, providing greater effectiveness than an equal number ofvehicles,operating ,independently. ,Collaborative Autonomy,allows the human ,warfighter to command ,the unmanned,vehicles as an ,active member ,of a ,warfighting team, rather than as a detached controller (Figure 1). Central to the ,architecture are state-of-the-art software
How can the semantic interpretation of a formal symbol system be made
intrinsic to the system, rather than just parasitic on the meanings in our
heads? How can the meanings of the meaningless symbol tokens, manipulated
solely on the basis of their (arbitrary) shapes, be grounded in anything but
other meaningless symbols? The problem is analogous to trying to learn Chinese
from a Chinese/Chinese dictionary alone. A candidate solution is sketched:
Symbolic representations must be grounded bottom-up in nonsymbolic
representations of two kinds: (1) "iconic representations," which are analogs
of the proximal sensory projections of distal objects and events, and (2)
"categorical representations," which are learned and innate feature-detectors
that pick out the invariant features of object and event categories from their
sensory projections. Elementary symbols are the names of these object and event
categories, assigned on the basis of their (nonsymbolic) categorical
representations. Higher-order (3) "symbolic representations," grounded in these
elementary symbols, consist of symbol strings describing category membership
relations (e.g., "An X is a Y that is Z").
A nonlinear control problem has been posed by Bupp et al. to provide a benchmark for evaluating various nonlinear control design techniques. In this paper, the capabilities of the state-dependent Riccati equation (SDRE) technique are illustrated in producing two control designs for the benchmark problem. The SDRE technique represents a systematic way of designing nonlinear regulators. The design procedure consists of first using direct parameterization to bring the nonlinear system to a linear structure having state-dependent coefficients (SDC). A state-dependent Riccati equation is then solved at each point x along the trajectory to obtain a nonlinear feedback controller of the form u=−R-1(x)BT(x)P(x)x, where P(x) is the solution of the SDRE. Analysis of the first design shows that in the absence of disturbances and uncertainties, the SDRE nonlinear feedback solution compares very favorably to the optimal open-loop solution of the posed nonlinear regulator problem, the latter being obtained via numerical optimization. It is also shown via simulation that the closed-loop system has stability robustness against parametric variations and attenuates sinusoidal disturbances. In the second design it is demonstrated how a hard bound can be imposed on the control magnitude to avoid actuator saturation. © 1998 John Wiley & Sons, Ltd.
Our paper on the use of heuristic information in graph searching defined a path-finding algorithm, A*, and proved that it had two important properties. In the notation of the paper, we proved that if the heuristic function ñ (n) is a lower bound on the true minimal cost from node n to a goal node, then A* is admissible; i.e., it would find a minimal cost path if any path to a goal node existed. Further, we proved that if the heuristic function also satisfied something called the consistency assumption, then A* was optimal; i.e., it expanded no more nodes than any other admissible algorithm A no more informed than A*. These results were summarized in a book by one of us.
Recently, there has been growing interest in developing unmanned aircraft systems (UAS) with advanced onboard autonomous capabilities. This paper describes the current state of the art in autonomous rotorcraft UAS (RUAS) and provides a detailed literature review of the last two decades of active research on RUAS. Three functional technology areas are identified as the core components of an autonomous RUAS. Guidance, navigation, and control (GNC) have received much attention from the research community, and have dominated the UAS literature from the nineties until now. This paper first presents the main research groups involved in the development of GNC systems for RUAS. Then it describes the development of a framework that provides standard definitions and metrics characterizing and measuring the autonomy level of a RUAS using GNC aspects. This framework is intended to facilitate the understanding and the organization of this survey paper, but it can also serve as a common reference for the UAS community. The main objective of this paper is to present a comprehensive survey of RUAS research that captures all seminal works and milestones in each GNC area, with a particular focus on practical methods and technologies that have been demonstrated in flight tests. These algorithms and systems have been classified into different categories and classes based on the autonomy level they provide and the algorithmic approach used. Finally, the paper discusses the RUAS literature in general and highlights challenges that need to be addressed in developing autonomous systems for unmanned rotorcraft. © 2012 Wiley Periodicals, Inc.
Since the mid-90's, State-Dependent Riccati Equation (SDRE) strategies have emerged as general design methods that provide a systematic and effective means of designing nonlinear controllers, observers, and filters. These methods overcome many of the difficulties and shortcomings of existing methodologies, and deliver computationally simple algorithms that have been highly effective in a variety of practical and meaningful applications. In a special session at the 17th IFAC Symposium on Automatic Control in Aerospace 2007, theoreticians and practitioners in this area of research were brought together to discuss and present SDRE-based design methodologies as well as review the supporting theory. It became evident that the number of successful simulation, experimental and practical real-world applications of SDRE control have outpaced the available theoretical results. This paper reviews the theory developed to date on SDRE nonlinear regulation for solving nonlinear optimal control problems, and discusses issues that are still open for investigation. Existence of solutions as well as stability and optimality properties associated with SDRE controllers are the main contribution in the paper. The capabilities, design flexibility and art of systematically carrying out an effective SDRE design are also emphasized.
We consider the problem of Simultaneous Localization and Mapping (SLAM) from a Bayesian point of view using the Rao-Blackwellised Particle Filter (RBPF). We focus on the class of indoor mobile robots equipped with only a stereo vision sensor. Our goal is to construct dense metric maps of natural 3D point landmarks for large cyclic environments in the absence of accurate landmark position measurements and reliable motion estimates. Landmark es-timates are derived from stereo vision and motion estimates are based on visual odometry. We distinguish between landmarks using the Scale Invariant Feature Transform (SIFT). Our work defers from current popular approaches that rely on reliable motion models derived from odometric hardware and accurate landmark measurements obtained with laser sensors. We present results that show that our model is a successful approach for vision-based SLAM, even in large environments. We validate our approach experimentally, producing the largest and most accurate vision-based map to date, while we identify the areas where future research should focus in order to further increase its accuracy and scalability to significantly larger environments.
Miniature Unmanned Aerial Vehicles (UAVs) are currently being researched for a wide range of tasks, including search and rescue,
surveillance, reconnaissance, traffic monitoring, fire detection, pipe and electrical line inspection, and border patrol to
name only a few of the application domains. Although small/miniature UAVs, including both Vertical Takeoff and Landing (VTOL)
vehicles and small helicopters, have shown great potential in both civilian and military domains, including research and development,
integration, prototyping, and field testing, these unmanned systems/vehicles are limited to only a handful of laboratories.
This lack of development is due to both the extensive time and cost required to design, integrate and test a fully operational
prototype as well as the shortcomings of published materials to fully describe how to design and build a “complete” and “operational”
prototype system. This work attempts to overcome existing barriers and limitations by detailing the technical aspects of a
small UAV helicopter designed specifically as a testbed vehicle. This design aims to provide a general framework that will
not only allow researchers the ability to supplement the system with new technologies but will also allow researchers to add
innovation to the vehicle itself.
The state dependent Riccati equation was originally developed for the continuous time systems. In the paper the optimality of a discrete time version of the state dependent Riccati equation is considered. The derivation of the optimal control strategy is based on the Hamiltonian optimal solution for the nonlinear optimal control problem. The new form of the discrete state dependent Riccati equation with a correction tensor is derived. The prediction of the future trajectory is used in the derivation.
A globally consistent solution to the simultaneous localization and mapping (SLAM) problem in 2D with three degrees of freedom (DoF) poses was presented by Lu and Milios [F. Lu, E. Milios, Globally consistent range scan alignment for environment mapping, Autonomous Robots 4 (April) (1997) 333–349]. To create maps suitable for natural environments it is however necessary to consider the 6DoF pose case, namely the three Cartesian coordinates and the roll, pitch and yaw angles. This article describes the extension of the proposed algorithm to deal with these additional DoFs and the resulting non-linearities. Simplifications using Taylor expansion and Cholesky decomposition yield a fast application that handles the massive amount of 3D data and the computational requirements due to the 6DoF. Our experiments demonstrate the functionality of estimating the exact poses and their covariances in all 6DoF, leading to a globally consistent map. The correspondences between scans are found automatically by use of a simple distance heuristic.
Agents operating in the real world often have limited time available for planning their next actions. Producing optimal plans is infeasible in these scenarios. Instead, agents must be satisfied with the best plans they can generate within the time available. One class of planners well-suited to this task are anytime planners, which quickly find an initial, highly suboptimal plan, and then improve this plan until time runs out.A second challenge associated with planning in the real world is that models are usually imperfect and environments are often dynamic. Thus, agents need to update their models and consequently plans over time. Incremental planners, which make use of the results of previous planning efforts to generate a new plan, can substantially speed up each planning episode in such cases.In this paper, we present an A∗-based anytime search algorithm that produces significantly better solutions than current approaches, while also providing suboptimality bounds on the quality of the solution at any point in time. We also present an extension of this algorithm that is both anytime and incremental. This extension improves its current solution while deliberation time allows and is able to incrementally repair its solution when changes to the world model occur. We provide a number of theoretical and experimental results and demonstrate the effectiveness of the approaches in a robot navigation domain involving two physical systems. We believe that the simplicity, theoretical properties, and generality of the presented methods make them well suited to a range of search problems involving dynamic graphs.
Optimal heuristic searches such as A* search are widely used for planning but can rarely scale to large complex problems. The suboptimal versions of heuristic searches such as weighted A* search can often scale to much larger planning problems by trading off the quality of the solution for efficiency. They do so by relying more on the ability of the heuristic function to guide them well towards the goal. For complex planning problems, however, the heuristic function may often guide the search into a large local minimum and make the search examine most of the states in the minimum before proceeding. In this paper, we propose a novel heuristic search, called R* search, which depends much less on the quality of the heuristic function. The search avoids local minima by solving the whole planning problem with a series of short-range and easy-lo-solve searches, each guided by the heuristic function towards a randomly chosen goal. In addition, R* scales much better in terms of memory because it can discard a search state-space after each of its searches. On the theoretical side, we derive probabilistic guarantees on the sub-optimality of the solution returned by R*. On the experimental side, we show that R* can scale to large complex problems. Copyright © 2008, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.
This paper presents a heterogeneous, asynchronous architecture for controlling autonomous mobile robots which is capable of controlling a robot performing multiple tasks in real time in noisy, unpredictable environments. The architecture produces behavior which is reliable, task-directed (and taskable), and reactive to contingencies. Experiments on real and simulated real- world robots are described. The architecture smoothly integrates planning and reacting by performing these two functions asynchronously using heterogeneous architectural elements, and using the results of planning to guide the robot's actions but not to control them directly. The architecture can thus be viewed as a concrete implementation of Agre and Chapman's plans-as- communications theory. The central result of this work is to show that completely unmodified classical AI programming methodologies using centralized world models can be usefully incorporated into real-world embedded reactive systems.
We consider the problem of autonomously flying Miniature Aerial Vehicles (MAVs) in indoor environments such as home and office buildings. The primary long range sensor in these MAVs is a miniature camera. While previous approaches first try to build a D model in order to do planning and control, our method neither attempts to build nor requires a 3D model. Instead, our method first classifies the type of indoor environment the MAV is in, and then uses vision algorithms based on perspective cues to estimate the desired direction to fly. We test our method on two MAV platforms: a co-axial miniature helicopter and a toy quadrotor. Our experiments show that our vision algorithms are quite reliable, and they enable our MAVs to fly in a variety of corridors and staircases.
Motor schemas are proposed as a basic unit of behavior specification for the navigation of a mobile robot. These are multiple concurrent processes which operate in conjunction with associated perceptual schemas and contribute independently to the overall concerted action of the vehicle. The motivation behind the use of schemas for this domain is drawn from neuroscientific, psychological and robotic sources. A variant of the potential field method is used to produce the appropriate velocity and steering commands for the robot. An implementation strategy based on available tools at UMASS is described. Simulation results show the feasibility of this approach.
Since it was introduced to the English-speaking world in 1962, Lev Vygotsky's Thought and Language has become recognized as a classic foundational work of cognitive science. Its 1962 English translation must certainly be considered one of the most important and influential books ever published by the MIT Press. In this highly original exploration of human mental development, Vygotsky analyzes the relationship between words and consciousness, arguing that speech is social in its origins and that only as children develop does it become internalized verbal thought. In 1986, the MIT Press published a new edition of the original translation by Eugenia Hanfmann and Gertrude Vakar, edited by Vygotsky scholar Alex Kozulin, that restored the work's complete text and added materials to help readers better understand Vygotsky's thought. Kozulin also contributed an introductory essay that offered new insight into Vygotsky's life, intellectual milieu, and research methods. This expanded edition offers Vygotsky's text, Kozulin's essay, a subject index, and a new foreword by Kozulin that maps the ever-growing influence of Vygotsky's ideas.
Prior to the twentieth century, theories of knowledge were inherently perceptual. Since then, developments in logic, statistics, and programming languages have inspired amodal theories that rest on principles fundamentally different from those underlying perception. In addition, perceptual approaches have become widely viewed as untenable because they are assumed to implement recording systems, not conceptual systems. A perceptual theory of knowledge is developed here in the context of current cognitive science and neuroscience. During perceptual experience, association areas in the brain capture bottom-up patterns of activation in sensory-motor areas. Later, in a top-down manner, association areas partially reactivate sensory-motor areas to implement perceptual symbols. The storage and reactivation of perceptual symbols operates at the level of perceptual components--not at the level of holistic perceptual experiences. Through the use of selective attention, schematic representations of perceptual components are extracted from experience and stored in memory (e.g., individual memories of green, purr, hot). As memories of the same component become organized around a common frame, they implement a simulator that produces limitless simulations of the component (e.g., simulations of purr). Not only do such simulators develop for aspects of sensory experience, they also develop for aspects of proprioception (e.g., lift, run) and introspection (e.g., compare, memory, happy, hungry). Once established, these simulators implement a basic conceptual system that represents types, supports categorization, and produces categorical inferences. These simulators further support productivity, propositions, and abstract concepts, thereby implementing a fully functional conceptual system. Productivity results from integrating simulators combinatorially and recursively to produce complex simulations. Propositions result from binding simulators to perceived individuals to represent type-token relations. Abstract concepts are grounded in complex simulations of combined physical and introspective events. Thus, a perceptual theory of knowledge can implement a fully functional conceptual system while avoiding problems associated with amodal symbol systems. Implications for cognition, neuroscience, evolution, development, and artificial intelligence are explored.
We consider the problem of estimating detailed 3D structure from a single still image of an unstructured environment. Our goal is to create 3D models which are both quantitatively accurate as well as visually pleasing. For each small homogeneous patch in the image, we use a Markov random field (MRF) to infer a set of "plane parameters" that capture both the 3D location and 3D orientation of the patch. The MRF, trained via supervised learning, models both image depth cues as well as the relationships between different parts of the image. Inference in our model is tractable, and requires only solving a convex optimization problem. Other than assuming that the environment is made up of a number of small planes, our model makes no explicit assumptions about the structure of the scene; this enables the algorithm to capture much more detailed 3D structure than does prior art (such as Saxena et ah, 2005, Delage et ah, 2005, and Hoiem et el, 2005), and also give a much richer experience in the 3D flythroughs created using image-based rendering, even for scenes with significant non-vertical structure. Using this approach, we have created qualitatively correct 3D models for 64.9% of 588 images downloaded from the Internet, as compared to Hoiem et al.'s performance of 33.1%. Further, our models are quantitatively more accurate than either Saxena et al. or Hoiem et al.