It’s a challenge we see all the time: a small municipality with a great vision for an abandoned industrial site, a tight timeline, and a very limited budget. How do you help them actually see their potential without breaking the bank?
Recently, we tackled exactly this. We needed to visualize a development that would link an existing public park to a downtown area. Instead of jumping straight into expensive, time-consuming renders, we broke it down into a 3-part scenario planning series.
Here is how we did it:
We began with the absolute basics—Level of Detail-1. Using basic 3D building masses, we mapped out the spatial foundation.
Why start so stripped-down? Because it focuses the conversation. When you show stakeholders simple blocks, they talk about what actually matters at this stage: scale, density, and flow, rather than getting distracted by subjective architectural tastes. Plus, simple 3D envelopes prove that the design is still open to their feedback. Hyper-realistic renders early on can accidentally look too "finished" and discourage people from sharing their thoughts.
Once the layout was approved, we needed to make it feel real. Hand-modeling every building costs a fortune, so we leaned heavily on procedural generation using Esri CityEngine.
By writing specific CGA rules, we procedurally generated building facades and automated streetscapes that reflected the actual character of their downtown. This mid-level detail is the ultimate sweet spot.
The community recognizes their town, making the changes feel tangible, but because it’s driven by code, we can still adjust heights, setbacks, or land uses with a single click if the stakeholders change their minds.
Moving these procedural models into Twinmotion gives us that beautiful, immersive environmental context. But we wanted to push it further.
Inspired by Esri’s AI Screenshot Stylizer for CityEngine, we ran our urban massings through a text-to-image AI workflow. In just minutes, we were swapping aesthetics—jumping from Modern Minimalist to Industrial Brutalist to Bauhaus. What traditionally takes weeks of heavy rendering now takes a few minutes of rapid-fire creative exploration. It allows us to test wild ideas long before committing to a final look.
The Big Takeaway High-quality GeoDesign isn't just for mega-cities anymore.
If you stagger your approach (starting at LOD 1, moving to LOD 2, and finishing with AI), you keep the community engaged with how a space functions before getting distracted by how it looks. It is hands-down the most efficient way to iterate quickly, keep costs down, and maximize stakeholder buy-in.