---
title: Reasoner–Planner Architecture
url: https://www.emergentmind.com/topics/reasoner-planner
type: topic
---

# Reasoner–Planner Architecture

A Reasoner–Planner is a composite architectural framework for intelligent agents, in which a "reasoner" module produces inferences, beliefs, or abstract goals from a knowledge base or sensory input, and a "planner" module generates and/or executes action sequences to accomplish specified objectives, often under uncertainty, partial observability, or dynamic feedback. This paradigm underpins a wide range of systems in robotics, embodied AI, tool-augmented language models, fact-seeking QA, and multi-task logic environments. Hybrid designs exploit the complementary strengths of logic, probabilistic modeling, learning, and control. Core features include explicit separation between reasoning (often with logic-based or probabilistic rules) and planning (often with action selection, symbolic planning, or policy optimization), bidirectional feedback, and modules for explanation, adaptation, and error recovery [1508.00059, 1901.05322, 2505.07500, 2511.10037, 2402.00658, 2510.16767, 2410.08328, 2505.01479, 2512.03560, 2501.10106, 2601.08282, 2410.09252, 2409.18545, 2603.14458, 2503.08508, 2202.00531].

## 1. Architectural Principles and Canonical Forms

The Reasoner–Planner pattern involves (at minimum) two modules:

- **Reasoner:** Implements logic-based or probabilistic inference over explicit knowledge representations, transforming raw inputs and priors into beliefs, explanations, or abstract sub-goals.
- **Planner:** Generates, validates, or executes action sequences—often considering domain constraints, uncertainty, or user goals—using information provided by (and sometimes modifying) the reasoner’s output.

Key instantiations include:
- Mixed logical-probabilistic frameworks, with Answer Set Programming (ASP) for symbolic inference and planning, paired with Bayesian belief updates for uncertainty management [1508.00059].
- Hybrid decision-making pipelines combining logical inference (e.g. P-Log), learned classifiers, and probabilistic planning (e.g., POMDPs) [1901.05322].
- Large Language Model (LLM)-augmented modular systems splitting “goal reasoning” (extracting intent, subgoals, or tool structure) from “action planning” (step-wise action selection under complex multi-modal input) [2505.07500].
- Planner-centric architectures decomposing tool calling and workflow orchestration in LLM agents into global DAG planning and local execution [2511.10037].
- Deep neural reasoning modules paired with tree/tree-of-thought planning over symbolic world models and iterative correction loops [2402.00658, 2505.01479].
- Multi-agent, explicitly decoupled designs for strategic decomposition (Reasoner–Planner) supervised execution (ReAct), and pipeline context management [2512.03560].

## 2. Formal Foundations and Logic-Probabilistic Encoding

Formally, Reasoner–Planner systems encode world knowledge and planning problems using logic programs, epistemic models, or explicit PDDL-style symbolic state/action spaces:
- **ASP/CR-Prolog module:** Domain knowledge captured as causal laws, state constraints, executability conditions, defaults with exceptions, and historical observations. Candidate plans (as answer sets) are generated to satisfy constraints and goals. Explanatory reasoning (via consistency-restoring rules) accounts for exogenous events and partial observations [1508.00059].
- **Belief update/commitment:** The planner or executor maintains a local probabilistic or Bayesian belief over a relevant subset of fluents, updating beliefs based on sensor input using formulas such as:
  $$
  P(E_i \mid O_i) = \frac{P(O_i \mid E_i)P(E_i)}{P(O_i \mid E_i)P(E_i) + P(O_i \mid \neg E_i)P(\neg E_i)}
  $$
  Probabilities above a threshold cause facts to be committed back to the logical program, closing the perception-reasoning loop [1508.00059].
- **Symbolic planning/validation:** Actions are proposed as grounded instances, validated/executed via a world model (PDDL, STRIPS, temporal logic, or temporal knowledge graphs), and checked for pre/postcondition satisfaction [2505.01479, 2410.09252, 2510.16767].
- **Learning-based symbolic interfaces:** LLM-based planners use symbolic world representations (TKGs, memory streams, parseable JSON graphs) as the basis for formulating and validating hypothetical plans [2410.09252, 2402.00658, 2505.01479].

## 3. Execution Workflow and Closed-loop Adaptation

The typical Reasoner–Planner workflow features tight interleaving of inference, planning, execution, and explanation steps:

1. **Perception and initialization:** Initial facts (sensor input, user instructions) are encoded as logical facts or symbolic state variables.
2. **Reasoning phase:** The reasoner computes beliefs, infers subgoals, detects inconsistencies (unexpected outcomes or partial observations), and generates explanations using non-monotonic rules or epistemic models [1508.00059, 2409.18545, 2410.08328].
3. **Plan extraction:** The planner selects or synthesizes a candidate plan (sequence of actions or tool invocations) to satisfy given goals, accounting for current beliefs and constraints [1508.00059, 1901.05322, 2503.08508, 2501.10106].
4. **Action execution:** Each action is executed (physically, by tool call, or as a simulated step); relevant state or observation is updated.
5. **Belief update & feedback:** Sensor or tool feedback drives probabilistic belief update and may trigger replanning or explanation if discrepancies or partial scenes are detected [1508.00059, 2510.16767].
6. **Adaptation and correction:** Recovery from failure is handled by CR rules, plan revision (dynamic plan rewriting in retrieval-augmented LLMs), or counterexample-driven candidate regeneration (LLM and STL loops in T³ Planner) [1508.00059, 2510.16767, 2601.08282].
7. **Termination and reporting:** When the goal is achieved or the planner signals irrecoverable failure, the process ends; explanations for anomalies are provided in the logical output or via the communication channel [1508.00059, 2501.10106].

## 4. Representative Domains and Application Scenarios

Reasoner–Planner architectures have been successfully applied in:

- **Mobile and service robotics**: Restaurant waiter robot domains with complex task sequencing, exogenous disturbances, and perception-driven planning [1508.00059].
- **Human-robot interaction**: Epistemic planners that explicitly track and manipulate nested beliefs, using AND–OR search to synthesize when to communicate (ask/inform), defer action, or explain to a human collaborator [2409.18545].
- **Vision-and-language navigation**: Embodied agents parsing human instructions into symbolic subgoals and using hierarchy (LLM for goal, LoRA for low-level action) for traceable and robust navigation [2505.07500].
- **Task-level manipulation and real-world robotic control**: Lightweight LLM decision modules for parameterized motion skill selection, with explicit verification of semantic consistency and parameter validity [2503.08508].
- **Complex multi-tool LLM agents**: Planner-centric tool orchestration (global DAGs with node/edge optimization), supervisory Reasoner–Planner frameworks for strategic decomposition and efficient tool usage [2511.10037, 2512.03560].
- **Fact-seeking QA and retrieval-augmented generation**: Teacher-student distilled planners that output explicit stepwise decomposition and fact request graphs, decoupled from retrieval and answer synthesis [2603.14458, 2601.08282].
- **Multi-task and multi-relational logic environments**: Deep RL-driven selection of modular logic operators, balancing universal reasoning rules with efficient per-instance pruning [2202.00531].

## 5. Empirical Results and Comparative Performance

Empirical evaluations consistently demonstrate the effectiveness of Reasoner–Planner architectures:

| Setting                                 | Metric                       | Pure Logic | Pure Probabilistic | Reasoner–Planner (Hybrid) |
|------------------------------------------|------------------------------|------------|--------------------|---------------------------|
| Restaurant robot service [1508.00059]    | Task success rate            | 0.82       | 0.99               | 1.00                      |
|                                          | Completion time (normalized) | 1.06       | 3.32               | 1.00                      |
| Office SDM (LCORPP) [1901.05322]         | F₁ intention estimation      | 0.75(R+P)  | 0.70(P-only)        | 0.82                      |
| Fact QA (distilled planner) [2603.14458] | SEAL-0 QA accuracy (%)       | 1.8–6.3    | –                  | 10.8                      |
| VLN (PEAP-LLM) [2505.07500]              | SPL on val-unseen (%)        | 38.88      | –                  | 40.98                     |
| Symbolic plan validation [2505.01479]    | PlanBench overall acc. (%)   | 17.5       | –                  | 50.0                      |

Hybrid Reasoner–Planner approaches outperform single-paradigm systems both in success rate and explanation robustness, and are more sample efficient when faced with partial knowledge, sparse data, and uncertainty.

## 6. Limitations, Open Problems, and Research Directions

Limitations and trade-offs of current Reasoner–Planner systems include:

- **Boundary calibration:** Determining the optimal resolution boundary between symbolic (logical) and probabilistic (Bayesian or statistical) representations remains an open modeling challenge [1508.00059].
- **Planning complexity:** Worst-case symbolic planning (e.g., ASP) can be computationally expensive in large or high-horizon domains—requiring horizon bounding or action locality to retain tractability [1508.00059, 2410.09252].
- **Committing beliefs:** Committing probabilistically inferred facts to the logical model can introduce errors when belief thresholds are poorly chosen, though non-monotonic recovery (CR rules) offers resilience [1508.00059].
- **Scale and abstraction:** LLM planners, while more general, often require external memory, iterative correction (IC), or symbolic world models to remain efficient and verifiable for long-horizon or complex tasks [2505.01479, 2512.03560].
- **Feedback and self-correction:** Architectures with strong self-correction (counterexample-driven feedback, plan revision) show higher empirical robustness but can incur additional computational cost and require elaborate prompt engineering [2510.16767, 2601.08282].
- **Integration overhead:** Explicit decoupling can introduce latency for short tasks due to multi-agent or staged communication [2512.03560, 2511.10037].
- **Context maintenance:** For systems using language models or restricted context windows, careful management of action/observation history (summaries or memory modules) is essential for long-horizon planning [2503.08508, 2512.03560].
- **Evaluation coverage:** Performance remains upper-bounded by the quality of external retrievals, symbolic validation, or available sensor data [2603.14458, 1508.00059].

Key open research directions include adaptive symbolic–probabilistic splitting, online learning of threshold and verification strategies, hierarchical and multi-modal integration, and more efficient/expressive planning in open-world or fully interactive settings.

## 7. Comparative Analysis and Historical Context

Earlier architectures treated reasoning and planning as monolithic (fully intertwined in logic or logic+probability), but practical challenges in sensor-driven, tool-augmented, and language-model agents have driven increasing modularization:
- Clear separation between abstraction-levels (intent extraction vs. concrete action planning) correspond to layered architectures in robotics, multi-agent AI, and conversational agents [1508.00059, 2410.08328, 2409.18545].
- Modern systems employ programmatic interfaces between modules—using fact commitment, action parameter passing, or explicit fact-query/result channels—to interleave symbolic, probabilistic, and learning-based modules [2505.01479, 2512.03560, 2603.14458].
- In epistemic planners, explicit modeling of agent nested beliefs and perspective-taking gives rise to communication-optimized strategies grounded in formal logic [2409.18545].
- End-to-end differentiable joint training (deep RL over logic operators, as in PRIMA [2202.00531]) unifies logic deduction rule learning and dynamic inference-path optimization, achieving both high generalization and efficiency.

The Reasoner–Planner formalism has set a new standard for generalizability, sample efficiency, robustness to partial knowledge, and human-transparent explanations in modern AI systems. Its continued evolution is a focal point in research across robotics, dialogue agents, retrieval-augmented LLMs, and symbolic-neuro reasoning hybrids.

Source: https://www.emergentmind.com/topics/reasoner-planner