Rosenblatt, J. K. (1995). DAMN: A distributed architecture for mobile navigation. In Proc. of the AAAI Spring Symp. on Lessons Learned from Implememted Software Architectures for Physical Agents, pages 167--178, Stanford, CA.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....to continuously sample the low level configuration space. Other integrated architectures [2] 6] 9] 12] 14] utilize potential fields [11] in their reactive controllers to encode continuous responses, which are subject to local minima problems [13] In contrast, integrated architectures [24], 25] that employ discrete response encoding (i.e. finite, enumerated set of responses) encode high level motion commands (e.g. forward,left,right, etc. which may not be physically realizable due to negligence of kinematic constraints (e.g. non holonomy) Interpolation of these ....

....complex obstacles) requiring a multitude of concurrently active behaviors. C. Achieve asynchronous execution of behavioral modules The asynchronous execution of behaviors, each at its fastest rate possible, is crucial to the preservation of reactive capabilities. Integrated architectures (e.g. [24], 25] that coordinate their behaviors to operate directly on the action representation require them to be synchronized to produce a meaningful action. On the other hand, our architecture utilizes a time independent action selection module to learn a neural map representation of the local ....

J. K. Rosenblatt. DAMN: A distributed architecture for mobile navigation. Journal of Experimental and Theoretical Artificial Intelligence, 9(2-3):339--360, 1997.

....and extensible than centralized approaches. Arkin [1] also emphasized the importance of a nonhierarchical broadcast of information. Furthermore, we propose a model for cooperation and competition between activities based on a simple bidding mechanism. A similar model was proposed by Rosenblatt [16] in the CMU s DAMN project. A set of modules cooperated to control a robot s path by voting for various possible actions, and an arbiter decided which was the action to be performed. However the set of actions was pre defined, while in our system each agent can bid for any action it wants to ....

J. Rosenblatt. Damn: A distributed architecture for mobile navigation. In Proceedings of the 1995.

....ALGORITHM We want to produce a means by which an autonomous robot in possession of a feature based representation of its surroundings can decide where to move so as to best explore it environment. This is an action selection problem. There is a large literature on action literature [12] [13], 14] but little written from the perspective of map building in the presence of uncertainty the issue this paper addresses. We assume the vehicle is equipped with a SLAM algorithm that is capable of consistent mapping and localization. No restrictions whatsoever are placed on the form of this ....

Julio K. Rosenblatt, DAMN: A Distributed Architecture for Mobile Navigation, Ph.D. thesis, Carnegie Mellon University, Pittsburgh, PA, 1997.

....and a hierarchical architecture based on Real time Control Systems (RCS) is currently in use for the Demo III program. The main objective of the Demo III program is to create an autonomous scouting vehicle capable of traversing an unstructured off road environment. Behavior based systems like [5] [6], 7] 8] share many common components with hier archical architectures with some important differences. They integrate several goal oriented behaviors simultaneously. In most cases several behaviors are generated, and an arbiter or decision maker weighs these behav iors to create an ....

J. Rosenblatt, "DAMN: A distributed architecture for mo- bile navigation," in AAAI 1995.

....neural network to continuously sample the low level configuration space. Other integrated architectures ( 2] 9] utilize potential fields [8] in their reactive controllers to encode continuous responses, which are subject to local minima problems [10] In contrast, integrated architectures ([15], 16] that employ discrete response encoding (i.e. finite, enumerated set of responses) encode high level motion commands (e.g. forward, left, right, etc. which may not be physically realizable due to negligence of kinematic constraints (e.g. non holonomy) Interpolation of ....

....complex obstacles) requiring a multitude of concurrently active behaviors. C. Achieve asynchronous execution of behavioral modules The asynchronous execution of behaviors, each at its fastest rate possible, is crucial to the preservation of reactive capabilities. Integrated architectures (e.g. [15], 16] that coordinate their behaviors to operate directly on the action representation require them to be synchronized to produce a meaningful action. On the other hand, our architecture utilizes a time independent action selection module to learn a neural map representation of the local ....

J. K. Rosenblatt. DAMN: A distributed architecture for mobile navigation. J. Experimental and Theoretical AI, 9(2-3):339--360, 1997.

....allowed to change its value, for all others it s value is read only. This mechanism adds multi robot cooperation capabilities into the DD framework, fitting perfectly in the global context of agent technology [Age] in which we are doing this work. DD differs from most other BBS, e.g. Ark98,SKR95,Ros97] by using dynamical system theory for the definition and analysis of behaviors. Furthermore this structure supports a practical interface for the integration of a deliberative control layer, see Sec. 5, 6. The implementation, evaluation and analysis of DD behaviors for different types of robots ....

J. K. Rosenblatt. DAMN: A distributed architecture for mobile navigation. Journal of Experimental and Theoretical Artificial Intelligence, 9(2/3):339--360, 1997.

....action selection mechanism in the deliberative layer) Following [10] cooperative action selection requires mechanisms that achieve some sort of behavior (or command) fusion , integrating information from different sources to obtain the current action. Examples are voting mechanisms [12, 11], superposition techniques [1] fuzzy command fusion mechanisms [17] or multiple objective behavior coordination methods [15] Competitive mechanisms, on the other hand, require the selection of an action based on the result of some competition process among different components, possibly ....

J. Rosenblatt. DAMN: A Distributed Architecture for Mobile Navigation. PhD thesis, Carnegie Mellon University Robotics Institute, Pittsburgh, PA, 1997.

....It combines the planning module of a deliberative model and the behavioral based control of a reactive model. coding that can produce very low level velocity or torque control of motor and joint actuators to perform fine, smooth motion control. This is in contrast to behavior based architectures [7, 9, 24, 25] and hybrid deliberative reactive architectures [3, 10, 11, 21, 22, 26] that employ discrete response encoding (enumerated set of responses) and thus, do not emphasize the significance of low level control. Our framework differs from existing architectures in the important ways: 1) The planning ....

J. Rosenblatt. DAMN: A distributed architecture for mobile navigation. In Spring Symposium on Lessons Learned for Implemented Software Architectures for Physical Agents, pages 167--178, 1995.

....which makes low level control possible. Our architecture adopts another form of continuous response encoding that can produce very low level velocity or torque control of motor and joint actuators to perform fine, smooth motion control. This is in contrast to behavior based architectures [6, 16, 17] and hybrid deliberative reactive architectures [1, 3, 14] that employ discrete response encoding (finite, enumerated set of responses) and, thus, do not emphasize the significance of low level control. Our framework differs from existing architectures in two important ways: 1) The planning ....

J. Rosenblatt. DAMN: A distributed architecture for mobile navigation. In Spring Symposium on Lessons Learned for Implemented Software Architectures for Physical Agents, pages 167--178, 1995.

....software organization eases the task of software configuration. Each module adds a certain competence, but not all modules are required to run the robot. The idea of decentralized, distributed decision making has been at the core of research on behavior based robotics over the last decade [1, 14, 78], but here modules are typically much lower in complexity (e.g. simple finite state machines) 3 Mobile Robot Localization 3.1 The Localization Problem A prime example of probabilistic computing in Minerva is localization. Localization is the problem of determining a robot s pose from sensor ....

J. Rosenblatt. DAMN: A Distributed Architecture for Mobile Navigation. PhD thesis, Robotics Institute, Carnegie Mellon University, Pittsburgh, PA, January 1997. Technical Report CMU-RI-TR-97-01.

.... the robot may be forced to react immediately to unforeseen problems (very likely, given an incomplete world model) Consequently, current mobile robot architectures recognize that planning heavy approaches to real time problems in dynamic environments (such as driving in traffic) are infeasible [18, 84]. MonoSAPIENT recognizes the critical role of perception in tactical driving and makes reasonable assumptions about available information (See Section 4.4) By abandoning the hope of generating complete plans from its world model, the chosen architecture interleaves planning and execution (as ....

.... robot community has moved towards architectures that emphasize interleaved planning and execution [30, 29] where, rather than searching completely to the goal state, the robot generates a partial plan and begins execution, possibly re planning from the current state at every new time step [84]. Not only does this make real time planning tractable, but it also allows the reasoning system to integrate the latest information, without committing the robot to potentially incorrect plans. The following sections detail important aspects in the triggering and execution of the lane change. ....

[Article contains additional citation context not shown here]

J. Rosenblatt. DAMN: A Distributed Architecture for Mobile Navigation. PhD thesis, Carnegie Mellon University, 1996.

....and extensible than centralized approaches. Arkin [1] also emphasized the importance of a nonhierarchical broadcast of information. Furthermore, we propose a model for cooperation and competition between activities based on a simple bidding mechanism. A similar model was proposed by Rosenblatt [16] in the CMU s DAMN project. A set of modules cooperated to control a robot s path by voting for various possible actions, and an arbiter decided which was the action to be performed. However the set of actions was pre defined, while in our system each agent can bid for any action it wants to ....

J. Rosenblatt. Damn: A distributed architecture for mobile navigation. In Proceedings of the

....of mixing humans with robots. One approach is to directly incorporate humans into the design, i.e. treating human perception or decision making as a system element. DAMN, for example, is a behavior based architecture in which individual modules vote for and against a range of possible actions[Rosenblatt95]. Command arbitration allows modules as disparate as autonomous safety behaviors and teleoperation to coexist. Another approach is the use of prioritized control, in which operator commands may be overridden by autonomous modules. The best known example of this type is NASREM, which explicitly ....

Rosenblatt, J., "DAMN: A Distributed Architecture for Mobile Navigation", in "Proceedings of the

....broadcast of information is not hierarchical. This approach is easier to program and is more flexible and extensible than centralized approaches. We propose a model for cooperation and competition between activities based on a simple bidding mechanism. A similar model was proposed by Rosenblatt [14] in CMU s DAMN project, in which voting was used to coordinate a set of modules to control the robot. The map building approach we use is based on the work by Prescott [13] who proposed a network model that stores the spatial relationships among landmarks for robot navigation. By matching a ....

J. Rosenblatt. Damn: A distributed architecture for mobile navigation. In Proc. of

No context found.

Rosenblatt, J. K. (1995). DAMN: A distributed architecture for mobile navigation. In Proc. of the AAAI Spring Symp. on Lessons Learned from Implememted Software Architectures for Physical Agents, pages 167--178, Stanford, CA.

No context found.

Rosenblatt, J. K. (1995). DAMN: a distributed architecture for mobile navigation. In Proceedings of the AAAI Spring Symp. on Lessons Learned from Implemented Software Architectures for Physical Agents, Stanford, CA.

No context found.

Rosenblatt, J.K.: DAMN: a distributed architecture for mobile navigation. In Proceedings of the AAAI Spring Symp. on Lessons Learned from Implemented Software Architectures for Physical Agents (1995)

No context found.

J. K. Rosenblatt. DAMN: A distributed architecture for mobile navigation. Journal of Experimental and Theoretical Artificial Intelligence, 9(2-3):339--360, 1997.

No context found.

Rosenblatt, J., "DAMN: A Distributed Architecture for Mobile Navigation", Doctoral Dissertation, Technical Report CMU-RI-TR-97-01, Robotics Institute, Carnegie Mellon University, January, 1997.

No context found.

J. K. Rosenblatt. DAMN: A distributed architecture for mobile navigation. In Proc. of the AAAI Spring Symp. on Lessons Learned from Implememted Software Architectures for Physical Agents, Stanford, CA, 1995.

No context found.

J. Rosenblatt. DAMN: A Distributed Architecture for Mobile Navigation. PhD thesis, Carnegie Mellon University Robotics Institute, Pittsburgh, PA, 1997.

No context found.

J. Rosenblatt, "DAMN: a distributed architecture for mobile navigation, " in Proc. AAAI Spring Symp. Lessons Learned From Implemented Software Architectures for Physical Agents, Stanford, CA, 1995, pp. 167--178.

No context found.

J. Rosenblatt, DAMN: A Distributed Architecture for Mobile Navigation. PhD thesis, Carnegie Mellon University, 1996.

No context found.

J. K. Rosenblatt, "DAMN: A distributed architecture for mobile navigation, " J. Expt. Theor. Artif. Intell., vol. 9, no. 2-3, pp. 339--360, 1997.

No context found.

J. Rosenblatt. Damn: A distributed architecture for mobile navigation. In Proceedings of the 1995 AAAI Spring Symposium on Lessons Learned from Implemented Software Architectures for Physical Agents, Stanford, CA,

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC