Automated planning and scheduling¶

Computational construction and temporal-resource organization of action sequences that move an agent system from an initial state toward specified goals.

Version: v2 · 2026-09-19 · History
Domain-specific #: 3370
Domain group: Applied Sciences & Engineering
Origin domain: Computer Science & Software Engineering
Subdomain: Artificial Intelligence → Computer Science & Software Engineering
Aliases: AI planning

Core Idea¶

Classical deterministic planning, temporal and resource-constrained scheduling, contingent and probabilistic planning and online replanning use different state, observability and execution assumptions. A domain model declares states, actions, preconditions, effects, durations and resources; search or optimization finds a feasible policy or plan and scheduling assigns times while execution feedback can trigger revision. The abstraction is therefore identified by a declared carrier, a transformation or constraint over that carrier, and an invariant that tells an analyst whether the named structure is genuinely present.

How would you explain it like I'm…

The Step-by-Step Plan Maker

Planning is figuring out which steps to do, and in what order, to reach a goal, like knowing you put on socks before shoes. A computer can do this too: it knows what each step needs first and what each step changes, and it hunts for a list of steps that reaches the goal. Scheduling decides when each step happens. If something goes wrong along the way, it can make a new plan.

Computers That Plan and Schedule

Automated planning and scheduling is when a computer figures out what actions to take and when, to reach a goal. First someone describes the world to it: what situations are possible, what each action needs before it can happen, what it changes, how long it takes, and what tools or resources it uses. The computer then searches for a plan that works, and scheduling assigns times so things don't clash. While the plan is running, if something unexpected happens, the computer can change the plan. Some planners assume everything goes exactly as expected; others plan for surprises and things they can't see.

AI Planning and Scheduling

Automated planning and scheduling is the part of AI that computes plans or policies for an agent. It starts from a domain model: states, actions with preconditions and effects, costs, durations, resource limits, an initial state, and goals. Search or optimization finds a feasible plan (a sequence of actions) or policy (a rule for what to do in each situation), and scheduling assigns start times under time and resource constraints. Feedback during execution can trigger replanning. Importantly, it is a family with different assumptions: classical planning assumes a deterministic, fully known world, while contingent and probabilistic planning handle uncertainty and limited observation, and online replanning revises as it goes. To say a problem really is one of these, you must spell out the agent, state and action representation, goals, constraints, uncertainty, output type, optimization criterion, and how plans are checked and revised.

Automated planning and scheduling computes courses of action for an agent from a declarative domain model that specifies states, actions, preconditions, effects, costs, durations, and resources, together with an initial state and goals. Search or optimization produces a feasible plan (an action sequence) or policy (a state-to-action mapping), and scheduling assigns times subject to temporal and resource constraints; during execution, feedback can trigger plan repair or replanning. The field is not a single problem: classical deterministic planning, temporal and resource-constrained scheduling, contingent planning under partial observability, probabilistic planning, and online replanning rest on different assumptions about state, observability, and execution. A problem is properly identified only when the agent and environment, state/action representation, initial state and goals, preconditions/effects/costs, time and resource constraints, uncertainty and observability, output form, optimization criterion, and validation/replanning rule are explicit. Stating these prevents a method validated under one set of assumptions (say, determinism) from being transferred to a setting that violates them.

Scope of Application¶

Automated planning and scheduling belongs to artificial intelligence and is useful where the analyst can specify the typed artificial intelligence carrier, including objects, relations, parameters, conventions, evidence, boundaries, and comparison targets, then evaluate the agent and environment, state and action representation, initial state and goals, preconditions effects and costs, time and resource constraints, uncertainty and observability, plan or policy output, optimization criterion, validation and replanning rule are explicit. The scope is broad within that domain but bounded by the need for the agent and environment, state and action representation, initial state and goals, preconditions effects and costs, time and resource constraints, uncertainty and observability, plan or policy output, optimization criterion, validation and replanning rule are explicit.

Clarity¶

The abstraction clarifies a crowded vocabulary by making the agent and environment, state and action representation, initial state and goals, preconditions effects and costs, time and resource constraints, uncertainty and observability, plan or policy output, optimization criterion, validation and replanning rule are explicit the center of the account. A claim should name the carrier, the governing operation or relation, the applicable assumptions, and the recognition test.

Manages Complexity¶

Without the abstraction, an analyst must reason directly over many local details: the carrier roles, admissibility assumptions, competing conventions, derived invariants, boundary cases, and proof or validation obligations specific to Automated planning and scheduling. Automated planning and scheduling compresses them into the roles in the structural signature. That compression permits comparison across instances without erasing the variables that determine validity. It also exposes which details may be varied safely and which are constitutive.

Abstract Reasoning¶

Identify the carrier. State what the elements, states, objects, or observations are: the typed artificial intelligence carrier, including objects, relations, parameters, conventions, evidence, boundaries, and comparison targets. Reject examples whose alleged carrier belongs to a different problem.
Lock the constitutive rule. Express the agent and environment, state and action representation, initial state and goals, preconditions effects and costs, time and resource constraints, uncertainty and observability, plan or policy output, optimization criterion, validation and replanning rule are explicit independently of one notation or implementation.

Knowledge Transfer¶

Knowledge transfers strongly among subfields of artificial intelligence because they reuse the typed artificial intelligence carrier, including objects, relations, parameters, conventions, evidence, boundaries, and comparison targets, A domain model declares states, actions, preconditions, effects, durations and resources; search or optimization finds a feasible policy or plan and scheduling assigns times while execution feedback can trigger revision., and type the carrier, state every parameter and convention in the definition, test that the agent and environment, state and action representation, initial state and goals, preconditions effects and costs, time and resource constraints, uncertainty and observability, plan or policy output, optimization criterion, validation and replanning rule are explicit, compare the nearest accepted identity, and report counterexamples, uncertainty, and limiting cases.

Relationships to Other Abstractions¶

Current abstraction Automated planning and scheduling Domain-specific

Parents (1) — outgoing typed relations

Automated planning and scheduling is a kind of Planning Prime

It remains its own entry because its identity is fixed by the agent and environment, state and action representation, initial state and goals, preconditions effects and costs, time and resource constraints, uncertainty and observability, plan or policy output, optimization criterion, validation and replanning rule.

Hierarchy path (1) — routes to 1 parentless root

Automated planning and scheduling → is a kind of Planning → is a kind of Foresight

Neighborhood in Abstraction Space¶

Automated planning and scheduling sits in a crowded region of the domain-specific corpus (10^th percentile for distinctiveness): several abstractions share nearly its structure, so a description that fits it tends to fit its neighbors too.

Family — Knowledge Representation & Information Systems (31 abstractions)

Nearest neighbors

Distributed artificial intelligence — 0.93
Model-based reasoning — 0.93
Constraint satisfaction — 0.92
Knowledge representation and reasoning — 0.92
Qualification problem — 0.91

Computed from structural-signature embeddings · 2026-10-25