AI has the potential to address challenges in adaptive reuse decision-making. As part of the Reincarnate project, we created desirable futures for the circular adaptive reuse of buildings through a series of scenarios. These scenarios contribute to Reincarnate’s broader goal of developing technical and social strategies to create opportunities for buildings, construction products, and materials.

The scenario development process included a novel approach that combined traditional scenario methodology (cross-impact balance analysis) with participatory scenario-planning workshops. We used AI tools to collaboratively develop narrative and visual elements with stakeholders, which allowed us to better understand their benefits and how they can accelerate adaptive reuse decision-making.

The outcomes include 15 detailed scenarios that can guide stakeholders in exploring future pathways for circular adaptive reuse, with practical implications for policy and project implementation. These are “big picture” scenarios for adaptive reuse that represent hypothetical possibilities and are not related to specific projects.

Through this approach, we were able to consider a wide range of parameters that would be difficult to combine and analyze in a handcrafted scenario, ensuring a more comprehensive, data-driven outcome. This resulted in scenario scorecards like the one shown in Figure 1.

For 15 descriptors (e.g., political and community support, cost), three potential variants were drawn up:

1. Strong — descriptor’s objective was reached (happy green emoticon)

2. Medium — descriptor’s objective was partially reached (neutral yellow emoticon)

3. Weak — descriptor’s objective was not reached (sad red emoticon)

The relationships between these descriptor variants were mapped during a workshop with adaptive reuse experts. An algorithm calculated consistent combinations of variants, which served as draft scenarios. For 15 consistent draft scenarios, ChatGPT-4 generated storylines based on the descriptions of the variants. DALL-E, an extension in ChatGPT-4, generated images from these textual descriptions.

The scenario storylines served as input, accompanied by the prompt: “Create an image of an adaptive reuse scenario based on the scenario storyline above.” DALL-E transformed the prompts into visual outputs, but refinement was needed to ensure the visualizations aligned more closely with specific expectations. Two images were generated: one from outside of the building and one from inside the building looking out.

The sections below describe the managerial implications of using AI tools in the early stages of adaptive reuse decision-making and how they can benefit practitioners.

Accelerating the Early Design Process

Using AI visualization tools early in the design process has profound managerial implications. The traditional design process is lengthy and predominantly bottom-up: architects develop proposals for stakeholders to review and comment on. This sequential approach leads to delays and limited stakeholder engagement in the initial stages. By integrating AI visualization tools, stakeholders can be involved at the outset, collaboratively visualizing potential futures alongside architects.

This shift enables faster, more inclusive prototyping, allowing for iterative cycles in which scenarios are quickly explored, refined, or excluded based on collective input. The design process becomes more dynamic and responsive, reducing the time needed to arrive at an initial draft.

Aligning Stakeholder Wants & Needs

AI visualization tools can help align stakeholder wants and needs while enhancing their participation. In traditional design, differing stakeholder objectives lead to conflicting demands, making it hard to reach a consensus. AI visualization allows stakeholders to collaboratively visualize potential pathways, integrating these contradictory objectives in real time.

This approach accelerates the decision-making process and helps stakeholders see the tangible impact of their preferences and compromises, fostering a more inclusive and participatory environment. When integrated outcomes can be viewed early in the design process, managers can quickly identify and address potential conflicts, ensuring that the final design better reflects the collective vision of all parties.

Envisioning the Future

Traditionally, skilled architects produce renderings that resemble the final outcome, a time-consuming process that requires specialized expertise. Moreover, capturing the parametric complexity of early design stages is usually challenging. AI visualization tools streamline this process by making abstract concepts more tangible, allowing stakeholders to quickly grasp the potential impact of a scenario and reducing the need for multiple iterations.

Faster exclusion of unviable options helps surface the contradictions inherent in the design process. By making these tensions more visible and quantifiable, AI helps overcome human biases that often cloud judgment during abstract design phases. It acts as a heuristic tool, guiding stakeholders toward more concrete and balanced decisions, leading to more informed and efficient outcomes. These benefits are particularly valuable in complex projects where competing objectives must be balanced. AI enables a clearer understanding of trade-offs, helping managers “put their finger” on the magnitude of the contradictions and address them quickly and effectively.

[For more from the authors on this topic, see: “Leveraging AI to Create a Circular Built Environment.”]