Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Large Language Model for Participatory Urban Planning

Zhilun Zhou, Yuming Lin, Depeng Jin, Yong Li

TL;DR

This paper presents an LLM-driven, multi-agent framework for participatory urban planning that simulates a planner and thousands of residents to generate land-use plans. A fishbowl discussion mechanism enables scalable resident deliberation across communities, with planner revisions guided by feedback. Evaluations on two Beijing regions show strong performance on need-aware metrics (Satisfaction, Inclusion) and competitive results on service and ecology, outperforming baselines and even human planners in some aspects. The approach highlights the potential and limitations of AI-assisted participatory planning for scalable, inclusive urban development.

Abstract

Participatory urban planning is the mainstream of modern urban planning that involves the active engagement of residents. However, the traditional participatory paradigm requires experienced planning experts and is often time-consuming and costly. Fortunately, the emerging Large Language Models (LLMs) have shown considerable ability to simulate human-like agents, which can be used to emulate the participatory process easily. In this work, we introduce an LLM-based multi-agent collaboration framework for participatory urban planning, which can generate land-use plans for urban regions considering the diverse needs of residents. Specifically, we construct LLM agents to simulate a planner and thousands of residents with diverse profiles and backgrounds. We first ask the planner to carry out an initial land-use plan. To deal with the different facilities needs of residents, we initiate a discussion among the residents in each community about the plan, where residents provide feedback based on their profiles. Furthermore, to improve the efficiency of discussion, we adopt a fishbowl discussion mechanism, where part of the residents discuss and the rest of them act as listeners in each round. Finally, we let the planner modify the plan based on residents' feedback. We deploy our method on two real-world regions in Beijing. Experiments show that our method achieves state-of-the-art performance in residents satisfaction and inclusion metrics, and also outperforms human experts in terms of service accessibility and ecology metrics.

Large Language Model for Participatory Urban Planning

TL;DR

This paper presents an LLM-driven, multi-agent framework for participatory urban planning that simulates a planner and thousands of residents to generate land-use plans. A fishbowl discussion mechanism enables scalable resident deliberation across communities, with planner revisions guided by feedback. Evaluations on two Beijing regions show strong performance on need-aware metrics (Satisfaction, Inclusion) and competitive results on service and ecology, outperforming baselines and even human planners in some aspects. The approach highlights the potential and limitations of AI-assisted participatory planning for scalable, inclusive urban development.

Abstract

Participatory urban planning is the mainstream of modern urban planning that involves the active engagement of residents. However, the traditional participatory paradigm requires experienced planning experts and is often time-consuming and costly. Fortunately, the emerging Large Language Models (LLMs) have shown considerable ability to simulate human-like agents, which can be used to emulate the participatory process easily. In this work, we introduce an LLM-based multi-agent collaboration framework for participatory urban planning, which can generate land-use plans for urban regions considering the diverse needs of residents. Specifically, we construct LLM agents to simulate a planner and thousands of residents with diverse profiles and backgrounds. We first ask the planner to carry out an initial land-use plan. To deal with the different facilities needs of residents, we initiate a discussion among the residents in each community about the plan, where residents provide feedback based on their profiles. Furthermore, to improve the efficiency of discussion, we adopt a fishbowl discussion mechanism, where part of the residents discuss and the rest of them act as listeners in each round. Finally, we let the planner modify the plan based on residents' feedback. We deploy our method on two real-world regions in Beijing. Experiments show that our method achieves state-of-the-art performance in residents satisfaction and inclusion metrics, and also outperforms human experts in terms of service accessibility and ecology metrics.
Paper Structure (24 sections, 12 figures, 2 tables, 1 algorithm)

This paper contains 24 sections, 12 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Participatory Planning
  4. Multi-agent Collaboration with Large Language Models
  5. Problem Statement
  6. Methods
  7. Agent Roles Design
  8. Planner Design
  9. Resident Design
  10. Planning Workflow
  11. Fishbowl Discussion
  12. Experiments
  13. Experiment Settings
  14. Datasets
  15. Implementation Details
  16. ...and 9 more sections

Figures (12)

  • Figure 1: The framework of our proposed participatory planning method.
  • Figure 2: The design of planner and residents.
  • Figure 3: The satellite image and area division of two regions, Huilongguan (HLG) and Dahongmen (DHM).
  • Figure 4: Performance comparison of our methods, our method without role-play, our method without discussion, and single planner.
  • Figure 5: (a) The discussion of residents and revision of plan made by the planner of community 1 in HLG. Changes in land-use are marked in bold solid lines in the map. (b) The metrics after the revision of each community. 0 indicates the initial plan, and 4 indicates the final plan after revising 4 communities. The metrics are the average of 5 runs.
  • ...and 7 more figures

Theorems & Definitions (1)

  • definition 1: Region Renovation