BDI Architecture
A classic agent design paradigm from the 1980s in AI and robotics, derived from Michael Bratman's theory of intentional action. It structures agents around three components—Belief, Desire, and Intention—and has had a significant influence on the design of autonomous agents.
Core Facts
The theoretical roots of multi-agent systems can be traced back to distributed artificial intelligence research in the 1980s.
The BDI (Belief-Desire-Intention) architecture originates from philosopher Michael Bratman's theory of intention and was a core framework in multi-agent systems research in the 1990s.
The BDI (Belief-Desire-Intention) model was proposed by Australian computer scientist Michael Bratman in 1987, and was later formalized into a computable framework by Rao and Georgeff.
Timeline (last 90 days)
The BDI architecture originates from the theory of rational agents in artificial intelligence and decomposes an agent's behavioral choices into three levels: beliefs, desires and intentions.
The SLA model relies on four core components: a belief-desire-intention (BDI) architecture, a precondition question graph, a dynamic interaction network, and a persistent publication registry.
The BDI (Belief-Desire-Intention) architecture is the most influential model in agent theory, holding that agents make rational decisions by maintaining beliefs, goals, and action plans.
The BDI (Belief-Desire-Intention) architecture from the 1990s had already implemented a similar perception-reasoning-action loop.
ROUNDS replaces the rule engine in traditional BDI systems with a large language model, making the agent's reasoning process more flexible and also less predictable.
ROUNDS replaces the hand-written rule engine relied upon by traditional BDI systems with a large language model, making agent reasoning more flexible and also more unpredictable.
The theoretical foundations of the Agent Loop can be traced back to the BDI (Belief-Desire-Intention) architecture of the 1980s and the agent-environment interaction loop of the Markov Decision Process (MDP) in reinforcement learning.
All Facts (12)
The theoretical roots of multi-agent systems can be traced back to distributed artificial intelligence research in the 1980s.
90%VerifiedThe BDI (Belief-Desire-Intention) architecture originates from philosopher Michael Bratman's theory of intention and was a core framework in multi-agent systems research in the 1990s.
80%VerifiedThe BDI (Belief-Desire-Intention) model was proposed by Australian computer scientist Michael Bratman in 1987, and was later formalized into a computable framework by Rao and Georgeff.
80%VerifiedThe BDI (Belief-Desire-Intention) model can be traced back to the 1990s and is a classic paradigm for Agent architecture in the field of artificial intelligence.
80%UnverifiedThe theoretical foundations of the Agent Loop can be traced back to the BDI (Belief-Desire-Intention) architecture of the 1980s and the agent-environment interaction loop of the Markov Decision Process (MDP) in reinforcement learning.
60%UnverifiedThe BDI (Belief-Desire-Intention) architecture was formally proposed by Rao and Georgeff in 1991, modeling agents as rational actors holding beliefs, desires, and intentions.
60%UnverifiedThe BDI architecture originates from the theory of rational agents in artificial intelligence and decomposes an agent's behavioral choices into three levels: beliefs, desires and intentions.
50%UnverifiedThe SLA model relies on four core components: a belief-desire-intention (BDI) architecture, a precondition question graph, a dynamic interaction network, and a persistent publication registry.
50%UnverifiedThe BDI (Belief-Desire-Intention) architecture is the most influential model in agent theory, holding that agents make rational decisions by maintaining beliefs, goals, and action plans.
50%UnverifiedThe BDI (Belief-Desire-Intention) architecture from the 1990s had already implemented a similar perception-reasoning-action loop.
50%UnverifiedROUNDS replaces the rule engine in traditional BDI systems with a large language model, making the agent's reasoning process more flexible and also less predictable.
50%UnverifiedROUNDS replaces the hand-written rule engine relied upon by traditional BDI systems with a large language model, making agent reasoning more flexible and also more unpredictable.
50%