2 Architectures Flashcards
3 architectures
- reactive architecture
- deliberative architecture
- hybrid architectures
reactive architecture
Focused on fast reactions to changes detected in the environment.
Every agent has very little intelligence, computing power and reasoning abilities. The union of a set of agents and the interaction between them allows the formation of highly complex, structured and efficient sytems.
reactive architecture
main characteristics
1 emergent functonality
- Simple agents
- Simple interaction
- Complex behaviour patterns appear as a result of the dynamic interactions
2 task decomposition
- agents composed of autonomous modules, each module manages a task
- minimal low level communication between modules
3 raw data
- basic data from sensors
- no complex symbolic data management
reactive architecture
advantages
- Simplicity of individual agents
- Flexibility, adaptability
- Ideal in very dynamic and unpredictable environments
- Low computational cost
- Avoiding complex planning/reasoning procedures
- Avoiding continuous model update
- Robustness against failure
- No central planning component
- Elegance
reactive architecture
limitations
- Agents without environment models must have sufficient information available from local environment.
- No long-term reasoning
- Limited applicability
- Difficult to make reactive agents that learn
- optimized to specific environments
- It is hard to engineer agents with large numbers of behaviours (dynamics of interactions become too complex to understand)
deliberative architecture
Focused on long-term planning of actions, centred on a set of basic goals.
- Explicit representation of the environment (map)
- Planning procedure that finds the minimal route between the current position and the destination
-> BDI model
deliberative architecture
main characteristics
- Explicit symbolic model of the world
- Decisions are made via logical reasoning, based on pattern matching and symbolic manipulation
- Sense-plan-act problem-solving paradigm of classical AI planning systems
deliberative architecture
limitations
- Dynamic world
- Update symbolic world model
- World changes while planning is being done
- Representation language
- Expressive enough to be useful in any domain
- Limited enough to be computationally tractable
- Classical planning. complete, optimal solutions
- High computational cost.
- Sometimes a sub-optimal low-cost fast reaction can be effective
hybrid architecture
Combining a reactive side and a deliberative side.
- a deliberative one, containing a symbolic world model, which develops plans and makes decisions in the way proposed by symbolic AI
- a reactive one, which is capable of reacting quickly to events without complex reasoning
- > Ferguson Touringmachines
- > Müller InteRRaP
deliberative architecture
limitations
- Lack of general methodologies to guide the design process
- Very specific, application dependent
BDI model
Belief - Desire - Intention
- beliefs: agents view of the environment
- desires: follow from beliefs
- goals: a subset of desires
- intentions: a subset of the goals
- plans: sequences of actions that are needed to archive the intentions, given its beliefs
Brooks
refutal of symbolic AI (key theses)
- intelligent behaviour does not require explicit representations
- intelligent behaviour does not require explicit abstract reasoning
- intelligence is an emergent property
key ideas:
- situatedness: real intelligence is in the world
- embodiment: real intelligence requires a physical body
- intelligence and emergence: intelligent behaviour arises out of interaction with its environment
Situated Automata
reactive
- agents are specified in a rule-like language
- this specifications is compiled to automation
- reasoning is done offline
- an agent is specified in terms of 2 componentes: perception (ruler) and action (gapps)
