Flying by Inference: Active Inference World Models for Adaptive UAV Swarms

arXiv 2026 47.4 method, application

TLDR

Expert-guided active inference framework for UAV swarm trajectory planning using learned probabilistic world models from offline demonstrations.

Reasoning

Strengths: Novel integration of active inference with expert demonstrations for hierarchical planning, validated with both simulations and real-flight data. Weaknesses: Limited comparison (only modified Q-learning), no explicit discussion of scalability or failure modes.

Read-first score

Read-first score 47.4, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 33.

Recency 6%

100

Uses a gentle age decay so recent papers surface without erasing older foundations. 2026

Citation impact 18%

66.5

Uses OpenAlex-shaped citation metadata as a bibliometric attention signal, separate from paper quality. citation_normalized_percentile=0.6648723

Methodology quality 18%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=result,validation

Topical relevance 29%

47.1

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Reproducibility 18%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=none

Citation velocity 12%

Citation velocity estimates citations per publication-year to reduce old-paper bias. velocity=0.00

Field roles

FrontierBridge

Rank sensitivity

Stability: volatile; rank range: 231.

Keyword Scores

world model

world dynamics prediction

generative world model

model-based reinforcement learning world model

interactive world model

world simulator

video world model

Deep Analysis

Innovations

Conversion of multi-UAV trajectory design from repeated combinatorial optimization into a hierarchical probabilistic inference problem
Expert-guided active-inference-inspired framework using genetic-algorithm planner with repulsive-force collision avoidance (GA-RF) to generate expert demonstrations
Learning of a probabilistic world model with Mission, Route, and Motion dictionaries that capture hierarchical dependencies
Online operation using posterior beliefs over symbolic states and KL-divergence-based abnormality indicators for mission allocation, route insertion, motion adaptation, and collision-aware replanning
Integration of Bayesian state estimators (EKF and PF) at the motion level for trajectory correction under uncertainty

Methodology

The method consists of an offline phase where a genetic-algorithm planner with repulsive-force collision avoidance (GA-RF) generates expert demonstrations, which are abstracted into Mission, Route, and Motion dictionaries. These dictionaries are used to learn a probabilistic world model that captures how expert mission allocations induce route orders and how route orders induce motion-level behaviors. During online operation, the UAV swarm evaluates candidate actions by forming posterior beliefs over symbolic states and minimizing KL-divergence-based abnormality indicators with respect to expert-derived reference distributions, enabling adaptive replanning without rerunning the offline optimizer.

Key Results

Simulation results show that the proposed framework preserves expert-like planning structure while producing smoother and more stable behavior than modified Q-learning. Additional validation using real-flight UAV trajectory data demonstrates that the learned world model can correct symbolic predictions under noisy and non-smooth observations.

Resources

Paper