MobEvolve: An Agentic Self-Evolving Heuristic System for Interpretable Human Mobility Generation
2026-06-01 • Artificial Intelligence
Artificial IntelligenceComputation and Language
AI summaryⓘ
The authors created MobEvolve, a new system that generates realistic travel patterns for people based on their traits. Unlike previous methods, MobEvolve starts with simple behavior rules and uses a language model to improve those rules step-by-step by learning from mistakes. Testing in Singapore and Montreal showed it makes more realistic and believable trip data while being easy to understand and fast to run compared to other advanced techniques. This helps better predict how people move without losing clarity or accuracy.
human mobility generationtrip chainsheuristic methodslarge language models (LLMs)behavioral plausibilitydistributional alignmentagent-based modelingtrajectory fidelityinference efficiency
Authors
Junlin He, Yihong Tang, Tong Nie, Ao Qu, Yuebing Liang, Hamzeh Alizadeh, Bang Liu, Wei Ma, Lijun Sun
Abstract
Human mobility generation aims to synthesize realistic trip chains for target populations based on individual features. Existing paradigms, including deep generative models, LLM-based methods, and traditional heuristics, struggle to satisfy the complex demands of this task while simultaneously maintaining interpretability, behavioral plausibility, population-level distributional alignment, and inference efficiency. To bridge this gap, we introduce MobEvolve, the first agentic self-evolving heuristic framework for human mobility generation. MobEvolve initializes a behavior-inspired heuristic system and employs an LLM agent to iteratively evolve its internal logic. By diagnosing empirical misalignments and failure cases on a validation set, the agent proposes targeted updates and accumulates evolution memory for cumulative self-improvement. Extensive evaluations on the Singapore and Montreal benchmarks demonstrate that MobEvolve significantly outperforms state-of-the-art deep generative and LLM-based methods in individual trajectory fidelity, population-level distribution alignment, and behavioral plausibility, while preserving interpretability and high inference efficiency.