Automated planning considering event based spatial object interactions using spatio-temporal reasoning

Knowledge Representation and Reasoning (KRR) in the domain of planning involve knowledge computing on expressiveness of action sequence that leads to desired goal state. An agent is an entity that perceives the environment. The essence of action sequence relies on representation of agent's perception towards the goal. The agent in an environment act on percept sequence to accomplish the states those lead to the goal state. Spatio-temporal reasoning based automated planning is proposed in which object interaction based on spatial relations in temporal scale is a state, representing an event and transition among the states are modeled using Finite State Machine (FSM) to extract the action sequence that accomplishes the task. Problem solving approach using Planning Domain Definition Language (PDDL) is significant when domain definitions involving agent's preconditions, actions and effects are required to model the percept sequence. PDDL is used to model agent's percept sequence with preconditions on actions and its effects. The development of proposed method into an autonomous planning agent with real time task orientation is left to work in future.