The situated approach in AI emphasizes building agents that act effectively within their environment through perception and action. It prioritizes grounded interaction and basic behavioral competence over abstract reasoning alone.
In artificial intelligence research, the situated approach builds agents that are designed to behave effectively successfully in their environment. This requires designing AI "from the bottom-up" by focussing on the basic perceptual and motor skills required to survive. The situated approach gives a much lower priority to abstract reasoning or problem-solving skills.