40:
336:
means of rules one can describe behaviour only prescriptively). However, the methods have also several flaws. First, for a designer, it is much more complicated to describe behaviour by a network comparing with if-then rules. Second, only relatively simple behaviour can be described, especially if adaptive feature is to be exploited.
308:. The conditions, states and actions are no more boolean or "yes/no" respectively but are approximate and smooth. Consequently, resulted behaviour will transition smoother, especially in the case of transitions between two tasks. However, evaluation of the fuzzy conditions is much slower than evaluation of their crisp counterparts.
173:. The meaning of the rule is as follows: if the condition holds, perform the action. The action can be either external (e.g., pick something up and move it), or internal (e.g., write a fact into the internal memory, or evaluate a new set of rules). Conditions are normally boolean and the action either can be performed, or not.
280:
step then. However, more often is the latter case. Here, every state is associated with a script, which describes a sequence of actions that the agent has to perform if its FSM is in a given state. If a transition activates a new state, the former script is simply interrupted, and the new one is started.
279:
There are two ways of how to produce behaviour by a FSM. They depend on what is associated with the states by a designer --- they can be either 'acts', or scripts. An 'act' is an atomic action that should be performed by the agent if its FSM is the given state. This action is performed in every time
262:
is only one of their possible applications. A typical FSM, when used for describing behaviour of an agent, consists of a set of states and transitions between these states. The transitions are actually condition action rules. In every instant, just one state of the FSM is active, and its transitions
335:
Positives of connectionist networks is, first, that the resulted behaviour is more smooth than behaviour produced by crisp if-then rules and FSMs, second, the networks are often adaptive, and third, mechanism of inhibition can be used and hence, behaviour can be also described proscriptively (by
331:
or free-flow hierarchies. The basic representational unit is a unit with several input links that feed the unit with "an abstract activity" and output links that propagate the activity to following units. Each unit itself works as the activity transducer. Typically, the units are connected in a
474:
196:, or may include special mechanisms for changing which goal / rule subset is currently most important. Flat structures are relatively easy to build, but allow only for description of simple behavior, or require immensely complicated conditions to compensate for the lacking structure.
283:
If a script is more complicated, it can be broken down to several scripts and a hierarchical FSM can be exploited. In such an automaton, every state can contain substates. Only the states at the atomic level are associated with a script (which is not complicated) or an atomic action.
375:, which map sensor inputs directly to effector outputs, and can follow or avoid. More complex systems are based on a superposition of attractive or repulsive forces that effect on the agent. This kind of steering is based on the original work on
351:: with a proper logic representation (which is suitable only for crisp rules), the rules need not to be re-evaluated at every time step. Instead, a form of a cache storing the evaluation from the previous step can be used.
203:
algorithms is a conflict resolution mechanism. This is a mechanism for resolving conflicts between actions proposed when more than one rules' condition holds in a given instant. The conflict can be solved for example by
287:
Computationally, hierarchical FSMs are equivalent to FSMs. That means that each hierarchical FSM can be converted to a classical FSM. However, hierarchical approaches facilitate designs better. See the
504:
467:
192:
which acts in response to an appropriate input. These layers are then organized into a simple stack, with higher layers subsuming the goals of the lower ones. Other systems may use
548:. Wortham, R. H., Gaudl, S. E. & Bryson, J. J., Instinct: A Biologically Inspired Reactive Planner for Intelligent Embedded Systems, In : Cognitive Systems Research. (2018)
69:
344:
Typical reactive planning algorithm just evaluates if-then rules or computes the state of a connectionist network. However, some algorithms have special features.
580:
276:. But transitions can also connect to the 'self' state in some systems, to allow execution of transition actions without actually changing the state.
543:. Platform for fast agent prototyping in Unreal Tournament 2004 – using POSH – reactive planner designed and developed by J.J. Bryson.
148:
There are several ways to represent a reactive plan. All require a basic representational unit and a means to compose these units into plans.
128:. Second, they compute just one next action in every instant, based on the current context. Reactive planners often (but not always) exploit
263:
are evaluated. If a transition is taken it activates another state. That means, in general transitions are the rules in the following form:
170:
421:
226:
121:
91:
413:
can be driven by this technique. In cases of more complicated terrain (e.g. a building), however, steering must be combined with
371:
Steering is a special reactive technique used in navigation of agents. The simplest form of reactive steering is employed in
477:. In: Johnson, W. L. (eds.): Proceedings of the First International Conference on Autonomous Agents. ACM press (1997) 22-29
355:
212:
52:
62:
56:
48:
446:
328:
244:
Conflict resolution is only necessary for rules that want to take mutually exclusive actions (c.f. Blumberg 1996).
460:
530:
van
Waveren, J. M. P.: The Quake III Arena Bot. Master thesis. Faculty ITS, University of Technology Delft (2001)
124:
in two aspects. First, they operate in a timely fashion and hence can cope with highly dynamic and unpredictable
484:. In: Proceedings of the Third International Conference on Autonomous Agents (Agents'99). Seattle (1999) 236-243
73:
488:
181:
241:
for selecting rules, but it is difficult to guarantee good behavior in a large system with simple approaches.
534:
105:
499:
495:
418:
354:
Scripting languages: Sometimes, the rules or FSMs are directly the primitives of an architecture (e.g. in
461:
Intelligence by Design: Principles of
Modularity and Coordination for Engineering Complex Adaptive Agents
258:(FSM) is model of behaviour of a system. FSMs are used widely in computer science. Modeling behaviour of
524:
414:
193:
410:
372:
255:
189:
176:
Production rules may be organized in relatively flat structures, but more often are organized into a
453:
289:
386:
359:
433:
293:
259:
125:
117:
200:
113:
28:
511:
514:. In: Computer Graphics International (CGI), IEEE Computer SocietyPress, New York (2005)
132:, which are stored structures describing the agent's priorities and behaviour. The term
348:
574:
324:
527:. Ph.D. Dissertation. Centre for Cognitive Science, University of Edinburgh (1993)
507:. In: Computer Graphics, 21(4) (SIGGRAPH '87 Conference Proceedings) (1987) 25-34.
481:
305:
247:
Some limitations of this kind of reactive planning can be found in Brom (2005).
17:
546:
406:
468:
AI Game
Development: Synthetic Creatures with learning and Reactive Behaviors
475:
Creatures: Artificial life autonomous software-agents for home entertainment
234:
177:
405:
The advantage of steering is that it is computationally very efficient. In
379:
of Craig
Reynolds. By means of steering, one can achieve a simple form of:
540:
498:. In Computational Intelligence, 23(4), 439–463, Blackwell-Wiley, (2005)
136:
goes back to at least 1988, and is synonymous with the more modern term
238:
362:, where the rules are only one of the primitives (like in JAM or ABL).
376:
219:
156:
A condition action rule, or if-then rule, is a rule in the form:
33:
449:. PhD thesis, Massachusetts Institute of Technology (1996).
27:"Dynamic planning" redirects here. For the anime studio, see
561:
517:
456:
In: Proceedings of MNAS workshop. Edinburgh, Scotland (2005)
512:
A motivational Model of Action
Selection for Virtual Humans
505:
Flocks, Herds, and
Schools: A Distributed Behavioral Model
463:. PhD thesis, Massachusetts Institute of Technology (2001)
447:
Old Tricks, New Dogs: Ethology and
Interactive Creatures
312:
358:). But more often, reactive plans are programmed in a
454:
Hierarchical
Reactive Planning: Where is its limit?
218:learning relative utilities between rules (e.g. in
208:
assigning fixed priorities to the rules in advance,
304:Both if-then rules and FSMs can be combined with
61:but its sources remain unclear because it lacks
292:of Damian Isla (2005) for an example of ASM of
482:JAM: A BDI-theoretic mobile agent architecture
525:Computational Mechanisms for Action Selection
8:
92:Learn how and when to remove this message
565:, an implementation of reactive planning
323:Reactive plans can be expressed also by
233:Expert systems often use other simpler
535:An Introduction to MultiAgent Systems
199:An important part of any distributed
184:consists of layers of interconnected
7:
491:. In: Gamastura online, 03/11 (2005)
473:Grand, S., Cliff, D., Malhotra, A.:
152:Condition-action rules (productions)
112:denotes a group of techniques for
25:
581:Automated planning and scheduling
496:Planning in Reactive Environment
313:architecture of Alex Champandard
296:, which uses hierarchical FSMs.
38:
211:assigning preferences (e.g. in
120:. These techniques differ from
537:. John Wiley & Sons (2009)
1:
489:Handling complexity in Halo 2
180:of some kind. For example,
340:Reactive planning algorithms
144:Reactive plan representation
597:
518:Softimage/Behavior product
510:de Sevin, E. Thalmann, D.:
494:Milani, A., Poggioni, V.,
392:a wall following behaviour
329:artificial neural networks
26:
383:towards a goal navigation
319:Connectionists approaches
169:. These rules are called
470:. New Riders, USA (2003)
182:subsumption architecture
47:This article includes a
106:artificial intelligence
76:more precise citations.
520:. Avid Technology Inc.
420:), which is a form of
325:connectionist networks
251:Finite State Machines
225:exploiting a form of
466:Champandard, A. J.:
373:Braitenberg vehicles
256:Finite state machine
190:finite state machine
417:(as e.g. in Milani
332:layered structure.
398:predator avoidance
387:obstacle avoidance
360:scripting language
294:computer game bots
274:activate-new-state
188:, each actually a
122:classical planning
49:list of references
434:Behavior based AI
395:enemy approaching
134:reactive planning
118:autonomous agents
110:reactive planning
102:
101:
94:
16:(Redirected from
588:
503:Reynolds, C. W.
300:Fuzzy approaches
201:action selection
138:dynamic planning
114:action selection
97:
90:
86:
83:
77:
72:this article by
63:inline citations
42:
41:
34:
29:Dynamic Planning
21:
18:Dynamic planning
596:
595:
591:
590:
589:
587:
586:
585:
571:
570:
567:by Grand et al.
558:
552:
533:Wooldridge, M.
442:
430:
401:crowd behaviour
369:
349:Rete evaluation
342:
321:
302:
253:
154:
146:
98:
87:
81:
78:
67:
53:related reading
43:
39:
32:
23:
22:
15:
12:
11:
5:
594:
592:
584:
583:
573:
572:
569:
568:
557:
556:External links
554:
550:
549:
544:
538:
531:
528:
521:
515:
508:
501:
492:
485:
480:Huber, M. J.:
478:
471:
464:
457:
450:
445:Blumberg, B.:
441:
438:
437:
436:
429:
426:
409:, hundreds of
407:computer games
403:
402:
399:
396:
393:
390:
384:
368:
365:
364:
363:
352:
341:
338:
320:
317:
301:
298:
252:
249:
231:
230:
223:
216:
215:architecture),
209:
153:
150:
145:
142:
130:reactive plans
100:
99:
57:external links
46:
44:
37:
24:
14:
13:
10:
9:
6:
4:
3:
2:
593:
582:
579:
578:
576:
566:
564:
560:
559:
555:
553:
547:
545:
542:
539:
536:
532:
529:
526:
523:Tyrrell, T.:
522:
519:
516:
513:
509:
506:
502:
500:
497:
493:
490:
486:
483:
479:
476:
472:
469:
465:
462:
458:
455:
451:
448:
444:
443:
439:
435:
432:
431:
427:
425:
423:
419:
416:
412:
408:
400:
397:
394:
391:
388:
385:
382:
381:
380:
378:
374:
366:
361:
357:
353:
350:
347:
346:
345:
339:
337:
333:
330:
326:
318:
316:
314:
309:
307:
299:
297:
295:
291:
285:
281:
277:
275:
272:
269:
266:
261:
257:
250:
248:
245:
242:
240:
236:
228:
224:
221:
217:
214:
210:
207:
206:
205:
202:
197:
195:
191:
187:
183:
179:
174:
172:
168:
165:
162:
159:
151:
149:
143:
141:
139:
135:
131:
127:
123:
119:
115:
111:
107:
96:
93:
85:
82:February 2011
75:
71:
65:
64:
58:
54:
50:
45:
36:
35:
30:
19:
562:
551:
459:Bryson, J.:
415:path-finding
404:
370:
343:
334:
322:
310:
303:
286:
282:
278:
273:
270:
267:
264:
254:
246:
243:
232:
198:
185:
175:
166:
163:
160:
157:
155:
147:
137:
133:
129:
126:environments
109:
103:
88:
79:
68:Please help
60:
306:fuzzy logic
171:productions
74:introducing
487:Isla, D.:
452:Brom, C.:
440:References
235:heuristics
563:Creatures
389:behaviour
268:condition
186:behaviors
178:hierarchy
161:condition
575:Category
541:Pogamut2
428:See also
422:planning
367:Steering
311:See the
237:such as
227:planning
239:recency
70:improve
260:agents
167:action
377:boids
327:like
290:paper
220:ACT-R
194:trees
55:, or
411:NPCs
356:Soar
271:then
213:Soar
164:then
116:by
104:In
577::
424:.
315:.
265:if
222:),
158:if
140:.
108:,
59:,
51:,
229:.
95:)
89:(
84:)
80:(
66:.
31:.
20:)
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.