Maes, P. 1992. Behavior-Based Artificial Intelligence. In Second Animat Conference on Adaptive Behavior, p. 15. Hawaii.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....decreases with the increase in the knowledge base size. Building a big centralized monolithic machine that has the capacity of doing many things might not be the best way to address a difficult problem. On the other hand, the decentralized (modular) way of thinking, or the bottom up approach ([14]) advocated by artificial life, could facilitate the construction of such a controller, leading to a behavior based architecture. So, keep the knowledge base modular and find a fuzzy controller for each behavior (or module of the application) can be seen as good heuristics for solving large ....

Pattie Maes. Behavior-based artificial intelligence. In Jean-Arcady Meyer, Herbert Roitblat, and Stewart Wilsom, editors, From Animals to Animats 2. MIT Press, 1993.

....4.3.4 or [Zhang and Muhlenbein, 1995] 8.1. 2 Adaptive representation ARL and architecture evolution The main critique to much of previous AI work for learning hierarchical decompositions is that decomposition architectures are hand designed in advance of solving the problem [Brooks, 1991; Maes, 1993b; Rumelhart et al. 1986; Ishikawa, 1995; Jacobs et al. 1991; Jordan and Jacobs, 1993] The ADF GP approach [Koza, 1992; Koza, 1994b] also starts with a generic architecture of a decomposition and modifies the components of the the decomposition throughout an evolutionary process, as opposed to a ....

Pattie Maes, "Behavior-Based Artificial Intelligence," In Proceedings of the Second Conference on Simulated and Adaptive Behavior. MIT Press, 1993.

....descriptions to the central planning and decision making modules. Moreover knowledge oriented theories do not include environmental pressures on the self preservation of the agent, and the role of adaptivity and emergence is taken over by the programmer. However, the claim (made for example in [66] that the classical, knowledge oriented approach works only for simulated toy problems and makes too many simplifying assumptions (e.g. static environments, single tasks, etc. is simply not true. Objective 10 results achieved in knowledge engineering for large scale, extremely challenging ....

....do at a particular moment, instead of making complex representations and following elaborated plans [47] It can be expected that many more design guidelines will become explicated as experience in building robotic agents continues. Some more extensive overviews can be found in [69] 22] 94] [66], a.o. 24 3.3 Different approaches are explored for designing the behavior programs. Although there seems to be a consensus in the field that behavior systems are appropriate units, different avenues are explored regarding the best way to design the underlying behavior programs. They fall ....

Maes, P. (1993) Behavior-Based Artificial Intelligence. In: Meyer, J-A., H.L. Roitblatt, and S.W. Wilson (1993) From Animals to Animats2. Proceedings of the Second International Conference on Simulation of Adaptive Behavior. MIT Press/Bradford Books. Cambridge Ma. p. 2-10.

....behavior is the ability to exploit diverse subsystems that embody a variety of methods. The ability to incorporate, use and maintain knowledge is but one possible method for selection of behavior as pointed out by Agre [4] and several methods have been used successfully on robots by others [5] [6], 7] Another important characteristic is the management of resources. For robots the resource management problem has typically been solved by explicit planning. However new action selection mechanisms provide the opportunity to combine resource management, focus of attention and intelligent ....

....intelligent action all under the umbrella of various action selection or arbitration mechanisms. A. Action Selection and Arbitration Viewing the robot as a whole, it operates by selecting actions for its actuators to achieve tasks and goals. Maes describes this as the action selection problem [6]. All architectures must provide a solution for this problem. Identifying the action selection problem places the focus on building up software that provides resources and mechanisms for action selection. The structure of any one approach and its emphasis may vary, but action selection or more ....

[Article contains additional citation context not shown here]

Pattie Maes, "Behavior-based artificial intelligence," in Proceedings of the 2nd Conference on Adaptive Behavior. 1993, MIT Press.

....interest in autonomous agents, are identified in section 2.3. 5 2.1 Behavior oriented AI vs. traditional AI Since 1985, a new approach to the study of Artificial Intelligence (AI) has emerged, perhaps best referred to as behavior oriented (e.g. Steels, 1994) or behavior based AI (e.g. Maes, 1993), as opposed to the traditional approach of knowledge based AI. Both AI approaches basically aim for the same general goal, the modeling and synthesis of intelligent behavior and cognitive capacities in artifacts. They do, however, disagree substantially in their notion of what intelligence ....

.... (e.g. Knight, 1990) Accordingly, knowledge based AI has traditionally emphasized the building of systems that possess a certain amount of knowledge about some problem domain (e.g. socalled expert systems ) that allows them to reason within this domain and, for example, answer a user s question (Maes, 1993). Hence, the central area of research within the field has been the problem of knowledge representation and knowledge engineering, i.e. 6 the problem of how to extract a human s (expert ) knowledge and represent it in a form appropriate for a machine to reason with. The knowledge based ....

[Article contains additional citation context not shown here]

Maes, P. (1993). Behavior-Based Artificial Intelligence. Proceedings of the Fifteenth Annual Meeting of the Cognitive Science Society (pp. 74-83). Hillsdale, NJ: Lawrence Erlbaum.

....use, and gives a thorough description of the various behaviors that are implemented within our robots. Section 4 describes and discusses experimental results that have been obtained. Section 5 concludes this paper. 2 AUTONOMOUS AGENTS Bottom Up AI [ 4 ] also referred to as Behavior based AI [ 16 ] in the literature) and Autonomous Agents consider intelligence as a biological feature (this notion is often referred to as Enactivism or Enaction [ 29 ] intelligence should be considered an agent s capacity to interact with its environment [ 22 ] rather than represent model it internally. ....

P. Maes. Behavior-Based Artificial Intelligence. In Proceedings of the Fifteenth Annual Meeting of the Cognitive Science Society, pages 74--83, Hillsdale, NJ, 1993. Lawrence Erlbaum.

....have seem qualitatively different from what is needed for low level parameter instantiation, and a separate conceptual level thus seems justified. 2.3 Relationship to Behavioral Robotics The control approach we describe falls under the general topic of behavioral robotics. In behavioral robotics [Maes, 1992; Brooks, 1986; Brooks, 1991] the robotic system is viewed as always being engaged in some particular task or behavior. Rather than looking for control formulations that will permit equally simple specification and execution of all actions the robot is physically capable of, the behavioral ....

Patty Maes, "Behavior-based Artificial Intelligence," in J. A. Meyer, H. Roitblat, and S. Wilson, editors, From Animals to Animats 2: Proceedings of the second International Conference of the simulation of Adaptive Behavior. Cambridge, MA: MIT Press, December 1992.

....oth er sensors step up power converter, 5V for CPU and memories step up power converter, 5V for other components BATTERY 1.2VNiCd cells voltage monitorin g circuit oscil ator 40 MHz Fig. 3: Hardware block diagram of SUBMAR II [17] The program structure is a behavior based (see [18] 19] and [20] for overview) Loosely coupled behaviors are running in parallel. The following generations of robots, intended to be used in particular industrial environments, will naturally be tailored for each case separately. Harsh environmental conditions will require a lot from the casing and the ....

Maes P: Behavior-Based Artificial Intelligence, In: From Animals to Animats 2: Proceedings of the Second Intl. Conf. on Simulation of Adaptive Behavior, ed. by Meyer J.-A., Roitblat H. L., Wilson S. W, The MIT Press, (1993), 2-10.

....is important from a scale up perspective. Ideally, a modular representation organizes the knowledge and competences of the agent such that local changes, improvements or tuning do not affect the functioning of most other components. Researchers in behavior based artificial intelligence (Maes 1993) talk about integrated competences or behaviors as given decompositions of the problem. We are interested in discovering decompositions that naturally emerge from the interaction agent environment. 1 For the same reason, parameterized function approximators have been used to replace table lookup ....

Maes, P. 1993. Behavior-based artificial intelligence.

....decisions in a complex and dynamic environment in order to achieve its autonomy. This is one example of the nouvelle AI and Artificial Life research the aim of which is to build autonomous system which can adapt to the world it is embedded in, this world being changing and often unpredictable [13] [20] To answer this new domain of robotics, Brooks has proposed a new designing methodology for mobile robots called subsumption architecture [2] His approach consists to build behavior based robots. To deal with the autonomous arbitration of the multiple behaviors, Brooks outlines an ....

Pattie Maes. Behavior-based artificial intelligence. In J-A. Meyer, H.L. Roitblat, and S.W. Wilson, editors, From Animals to Animats 2: Proceedings of the Second International Conference on Simulation of Adaptive Behavior, pages 2--10. MIT Press, 1993.

....of selforganising its own behaviour according to environmental constraints in order to maintain its own viability without human intervention. In recent years, the analogy between autonomous robots and biological organisms has generated a novel approach, also known as Behaviour Based Robotics [10], to program and understand robot behaviour in unknown and unpredictable environments. This approach is quite different from the classical AI approach that attempted to emulate human reasoning by building large and complex planning systems which failed to deliver the expected results [1] In ....

P. Maes. Behavior-based artificial intelligence. In J. Meyer, H. L. Roitblat, and S. W. Wilson, editors, From Animals to Animats II: Proceedings of the Second International Conference on Simulation of Adaptive Behavior. MIT Press-Bradford Books, Cambridge, MA, 1993.

....classic or centralized approach and the reactive approach ( 1] 2] They differ in the way each views the role played by a world model. The centralized or knowledge based approach emphasizes a (generally declarative) model of the world by representing detailed expertise from the problem domain [11]. It integrates sensing, reasoning or problem solving and acting modules [15] that operate on the domain knowledge. The reactive or behavior based approach recognizes the importance of the principles of connectedness and embodiment (the robots experience the world directly) Consequently, the ....

Pattie Maes. Behavior-based artificial intelligence. In Proceedings of the Second Conference on Simulated and Adaptibe Behavior. MIT Press, 1993.

....decisions in a complex and dynamic environment in order to achieve its autonomy. This is one example of the nouvelle AI and Artificial Life research the aim of which is to build autonomous system which can adapt to the world it is embedded in, this world being changing and often unpredictable [20] [29] To answer this new domain of robotics, Brooks has proposed a new designing methodology for mobile robots called subsumption architecture [3] His approach consists to build behavior based robots. To deal with the autonomous arbitration of the multiple behaviors, Brooks outlines an ....

Pattie Maes. Behavior-based artificial intelligence. In J-A. Meyer, H.L. Roitblat, and S.W. Wilson, editors, From Animals to Animats 2: Proceedings of the Second International Conference on Simulation of Adaptive Behavior, pages 2--10. MIT Press, 1993.

.... of expertise [32] 68] Most common drawbacks inherent to Classical AI are known as the Frame Problem (problem of maintaining a model of the real world [54] and the Grounding Problem (problem of relating the elements of the representation to the sensory information [30] Behavior based AI [41] (also referred to as Bottom Up AI, New AI or Autonomous Agents in the literature) on the other hand considers intelligence as a biological feature [45] this notion is often referred to as Enactivism or Enaction [67] intelligence should be considered an agent s capacity to interact with its ....

P. Maes. Behavior-Based Artificial Intelligence. In Proceedings of the Fifteenth Annual Meeting of the Cognitive Science Society, pages 74--83, Hillsdale, NJ, 1993. Lawrence Erlbaum.

....on a wrong place. The only possible solution is to provide the members a way to build some sort of primitive map of the environment and thus make simple planning possible. 5. Member s behavioral repertoire Both the real and simulated robots use behavior based control structures (see e.g. 22] [23] and [24] for overview) Used structures have similarities with both subsumption [25] and with Arkin s schema based ( 26] and [27] systems. Members have a special sensor block (i.e. perception behavior) which takes care of the interface to the sensors. The used sensors include: pressure, redox, ....

P. Maes, Behavior-Based Artificial Intelligence, From Animals to Animats 2 (2nd Intl Conf. on Simulation of Adaptive Behavior), J.-A. Meyer, H. L. Roitblat, S. W. Wilson (Eds.), MIT Press, Cambridge, MA, 1993, pp. 2-10.

....manner. The key to this successful line of research has been the development of the concept of behaviors. Behaviors Behaviors as a method of controlling robots were inspired by Brooks subsumption architecture (Brooks, 1986) and later generalized by Maes descriptions of behavior based designs (Maes, 1993). Generally, a behavior is a simple, focused, perceptual trigger and associated action. An example might look like: IF door on left THEN turn left. The door onleft is the perceptual trigger that activates the robot to turn left. The ideas of behaviors and fuzzy logic were a perfect match; ....

Maes, P. (1993). Behavior-Based Artificial Intelligence. In Proceedings of the Fifteenth Annual Conference of the Cognitive Science Society, 74-83, Boulder, CO.

....Agents constitute a whole discipline of Artificial Intelligence, whose description would be prohibitive to do here; as it is not the main concern of this paper, only the concepts of Autonomous Agents necessary to understand our implementation will be presented. More information can be found in [17] and [22] We will focus exclusively on autonomous agents that are considered to be embodied systems, which are designed to fulfill internal or external goals by their own actions in continuous long term interaction with the environment (possibly unpredictable and dynamical) in which they are ....

P. Maes. Behavior-Based Artificial Intelligence. In Proceedings of the Fifteenth Annual Meeting of the Cognitive Science Society, pages 74--83, Hillsdale, NJ, 1993. Lawrence Erlbaum.

....defined. It has been given many names by different researchers, including the animat approach [Wilson, 1985] reactive planning [Agre Chapman, 1987; Firby, 1987; Georgeff Lansky, 1987] computational neuroethology [Cliff, 1991] and the task oriented subsumption architecture [Brooks, 1986] Maes [1993] refers to these various approaches by a common underlying feature, calling it behavior based AI . Behavior based AI is a bottom up approach to autonomy that most often avoids declarative representation of domain knowledge. In place of knowledge structures and an inference engine are ....

....Washington, DC. Kaelbling, L. P. 1987) An architecture for intelligent reactive systems. In Reasoning about Actions and Plans, pages 395 410. Morgan Kaufmann, Los Altos, California. Lehnert, W. G. 1978) Representing physical objects in memory. Technical Report Research Report #131, Yale. Maes, P. 1990) Situated agents can have goals. In Maes, P. editor, Designing Autonomous Agents: Theory and Practice from Biology to Engineering and Back, pages 49 70. The MIT Press and Elsevier Science Publishers B.V. Cambridge, MA. Maes, P. 1993) Behavior based artificial intelligence. In Meyer, ....

[Article contains additional citation context not shown here]

Maes, P. (1993). Behavior-based artificial intelligence. In Meyer, J.-A., Roitblat, H., & Wilson, S. W., editors, From Animals to Animats: Proceedings of the Second International Conference on Simulation of Adaptive Behavior. The MIT Press, Cambridge, MA.

....manipulated. Brooks, 1991b, page 577] In traditional AI] the key issue on which emphasis is laid is a complete, correct internal model, a perfect copy of the world (with all its object and relationships) inside the system, which the system can rely on to predict how the problem can be solved. [Maes, 1993, page 4] Unfortunately, these claims are based on an incorrect view of the nature of the representational approach to AI. More specifically, they betray a lack of understanding of the nature of predicate logic, which underpins the symbolic paradigm. At a foundational level, the problem of ....

....page 570] Maes makes a similar point. In traditional AI] the central system evaluates the current situation (as represented in the internal model) and uses a search process to systematically explore the different ways in which this situation can be affected so as to achieve the desired goal. [Maes, 1993, page 4] Accordingly, a robot like Shakey is slow and intolerant to changes in its environment. This view is now uncontroversial, and robot architectures which incorporate a degree of reactivity are the norm. However, although systems which combine reactive and deliberative elements have been ....

P.Maes, Behavior-Based Artificial Intelligence, Proceedings of the Second International Conference on the Simulation of Adaptive Behavior (SAB 93), pages 2-10.

No context found.

Maes, P. 1992. Behavior-Based Artificial Intelligence. In Second Animat Conference on Adaptive Behavior, p. 15. Hawaii.

No context found.

Pattie Maes, "Behavior-Based Artificial Intelligence," In SAB-2. MIT Press, 1993.

No context found.

Maes, P. (1993). Behavior-Based Artificial Intelligence. In Proceedings of the Fifteenth Annual Conference of the Cognitive Science Society, pages 74-83, Boulder, CO.

No context found.

Maes, P. (1993). Behavior-Based Artificial Intelligence. In Proceedings of the Fifteenth Annual Conference of the Cognitive Science Society, pages 74-83, Boulder, CO.

No context found.

P. Maes. Behavior-based artificial intelligence. In [SAB92].

No context found.

Maes, P. "Behavior Based Artificial Intelligence". Proceedings of the Fifteenth Annual Conference of the Cognitive Science Society. 1993.

First 50 documents

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