The City Science Lab @ Shanghai is based in Tongji University. In the past few years, we have initiated projects including CityScope LivingLine, Equity WITHOUT Zoning and Sustainable Cities WITH Decentralization. This year, we integrate into SustainX and focus on the framework of hybrid community building, discussing the relationship between action and system, physical and virtual, our research focusing on data-driven human-surrounding interaction, networked collaborative intelligence, participatory design intervention, living ecosystem regeneration and beyond.
For this presentation we will host a series of lightning talks on the following projects:
Keynote: SustainX Schema: Design for Complex Socio-technical System
Prof. Yongqi LOU – Lab Director, Vice President of Tongji University, Professor
In 2014, during Tongji Design Week, a group of scholars including Don Norman, Ken Friedman, and Lou Yongqi unveiled the "DesignX Manifesto," which pinpointed design for "complex socio-technical systems" as a critical emerging direction in design research, education, and practice. To echo this new design culture, "Design and Data-Driven Hybrid Community Building" is strategically placed within a conceptual framework delineated by the "action-system" and "virtual-physical" axes. By fostering a hybrid community, we can create a series of unique incubators and living labs for responsible behavior and social interaction. Empowered by AI, Cloud technology and LLM, this Interactive Hybrid Community would act as a new public sphere, making it easier and more effective to drive large-scale, sustainable societal changes.
Project 1: Enhancing Knowledge-Driven Insights via IndustryScopeGPT
Leon Yang LIU - Lab Manager, Associate Professor
Siqi WANG - Ph.D. candidate
Industrial parks are critical to urban economic growth. Yet, their development often encounters challenges stemming from imbalances between industrial requirements and urban services, underscoring the need for strategic planning and operations. This presentation introduces IndustryScopeKG built upon the IndustryScope project, a pioneering large-scale multi-modal, multi-level industrial park knowledge graph, which integrates diverse urban data including street views, corporate, socio-economic, and geospatial information, capturing the complex relationships and semantics within industrial parks. Alongside this, we present the IndustryScopeGPT framework, which leverages Large Language Models (LLMs) with Monte Carlo Tree Search to enhance tool-augmented reasoning and decision-making in Industrial Park Planning and Operation (IPPO).
Our work significantly improves site recommendation and functional planning, demonstrating the potential of combining LLMs with structured datasets to advance industrial park management. This approach sets a new benchmark for intelligent IPPO research and lays a robust foundation for advancing urban industrial development.
Project 2: The Prosocial Urban Development Project
Ryan Yan ZHANG - Liaison, Affiliated Researcher
Chance Jiajie LI - Affiliated Researcher
Charlotte Jiwen GE - Affiliated Researcher
Brian Kejiang QIAN - Affiliated Researcher
Kai HU - Affiliated Researcher
Leon Yang LIU - Lab Manager, Associate Professor
The Prosocial Urban Development project introduces an innovative approach to urban planning, addressing challenges like housing affordability, congestion, and inequality. This research proposes an incentive-based land-use regulation system that aligns market forces with community demands to enhance urban social performance. The project balances interests of communities, developers, and planning authorities by promoting equitable distribution of ownership and decision-making rights. It incorporates a dual-option framework for developers, transparent voting processes, and community endowments. A key feature is an agent-based simulation powered by Large Language Models, predicting the system's urban and organizational performance. This simulation helps evaluate potential outcomes and refine the model. The project leverages blockchain technology, specifically Distributed Ledger Technologies and smart contracts, to ensure transparency and automate processes. By integrating community feedback and innovative incentives, this research aims to contribute to the ongoing dialogue on sustainable urban development and explore new possibilities for creating more responsive and inclusive urban environments.
Project 3: Spatial Intelligence Research for Adaptive Complex Community Scenarios
Tao SHEN – Associate Professor
This project explores the frontier of spatial intelligence within complex community environments. By harnessing advanced computational models and data analytics, we aim to develop innovative solutions that adapt dynamically to diverse urban scenarios. This research focuses on improving community planning, resource allocation, and resident interaction through intelligent, responsive systems designed to enhance the sustainability and livability of urban spaces.
Project 4: Intentional Community Building From Local to Global
Danwen ‘Eggy’ Ji – Ph.D Candidate
Kai HU - Affiliated Researcher
Unlike technology- or market-driven approaches, design-driven innovation is guided by a meaningful purpose and vision. Social innovation offers an experimental field where potential problems and solutions coexist.
This project explores design-driven social innovation, starting with the local NICE 2035 project and extending to global Pop-up Cities. By mapping and analyzing community dynamics through mixed methods—field studies, interviews, and network analysis—the research highlights three key aspects of community building for an envisioned future: building strong networks of trust and cultural exchange, developing scalable and context-specific solutions to address local issues, and empowering individuals to become community innovators and leaders.
Project 5: Multimodal Design Large Language Model and Innovative Applications
Meng WANG – Associate Professor
With the explosive development of large language model (LLM) technology, the design field is undergoing profound changes. However, tasks in the design field, such as sustainability, environment, digital, gaming, and UI/UX, require extensive domain-specific expert knowledge and data, which the current intelligence level of LLMs cannot satisfy. This report will introduce the collaboration between Tongji University and Tencent in the research and development of design LLM. It aims to develop a domain-specific LLM tailored for the design field, integrating design knowledge and advanced RAG technology, and exploring the opportunities and challenges of design large models from simple tools to complex systems.
Project 6: Power to People
Chang LIU - Associate Professor
Participatory design is a design method that is often used with community conversations for the design of neighborhood regeneration, social development, policy and other solutions. However, there are issues with the current participatory approach, raising concerns when it comes to the engagement levels, communication efficiency, and decision-making processes. This project explores the use of participatory approach for the design of AI technologies, and introduces a new approach that not only could be used for the real-time interactive visual feedback participatory approach, but it could bridge the gap between professionals and non-professionals in the community context. This is based on a research project conducted by a collaborative team consisting of the College of Design and Innovation, College of Electronic and Information Engineering at Tongji University, and Dell Technologies.
Project 7: A community Human and AI-robots Interaction
Jarmo SUOMINEN – Professor, Vice Lab Director
Kangyi ZHENG – Director of the Robotics Lab, Associate Professor
A community Human and AI-robots Interaction starts from the exhibition of NICE Commune in WDCC 2023. It is inspired by drawing on both sides of the exhibition wall. One is made by artist Jarmo Suominen, and the other is by robotic arm which generated by AI with Prompts.
We establish a comprehensive workflow for human-machine collaboration: By artificial intelligence data training, we trained a LoRA model in AI, based on the artistic expression of Jarmo’s works, and created its database. The public can generate digital images with the artist's style using prompts. Then automatically it converted into vector information by the computer and realized by a robotic arm simulating human drawing actions. Community can better unleash their imagination without being limited by their skills and the artist's personal capability is replicability and serviced to more places.
Recently, we fully realized the process in Huangpu high school. Empowering events and environments, we interviews with 12 stakeholders: cooks, cleaners, guards, teachers and students. Using data collection of their true stories, we transfer the Prompt engineering, generating visuals with Stable Diffusion using a robotic arm with AGV to draw the content. Finally artist Jarmo Suominen combines separate scenes into a complete mural.
This talk was presented by members of the City Science Lab @ Shanghai
(from top left to bottom right):
Prof. LOU Yongqi, Yang (Leon) LIU, Siqi WANG, Ryan Yan ZHANG, Chance Jiajie LI, Charlotte Jiwen GE, Brian Kejiang QIAN, Kai HU, Tao SHEN, Danwen ‘Eggy’ JI, Meng WANG, Chang LIU, Jarmo Suominen, and Kangyi ZHENG