From Ideas to Images: Exploring AI-Powered Visual Engagement

Some of the most valuable public engagement conversations are also among the hardest to bring online.

Think about a streetscape design workshop, a park planning charrette, or a neighborhood visioning session. Participants gather around a map, aerial image, or photograph and start imagining what could change: wider sidewalks, more trees, a protected bike lane, space for food trucks, a new playground, or an entirely different use for a public space.

In an in-person workshop, a skilled planner, landscape architect, or designer can quickly translate those ideas into sketches. With tracing paper, markers, and a practiced hand, concepts that begin as words can become something everyone around the table can see, discuss, and improve.

That process is powerful—but it is also difficult to scale. It often requires highly skilled graphic facilitators, works best with relatively small groups, and has traditionally relied on tools that don’t translate easily to online participation. Digital design software can help, but few people can operate tools like SketchUp quickly enough to support a live public conversation.

Recent advances in generative AI may offer another option.

At Konveio, we’ve been experimenting with a new prototype we currently call AI  Sketch, designed as an add-on to our Image Gallery activity.

The idea is simple: start with an existing image—a photograph of a street, for example, or a two-dimensional plan or map—and invite participants to describe what they would like to see.

A participant might ask for wider sidewalks, a row of street trees, bike infrastructure, outdoor seating, or food trucks. AI  Sketch then uses those instructions to generate a visual interpretation based on the original image, presented in an architectural sketch style.

Instead of requiring participants to know how to draw or use professional design software, their written ideas can become something visual that others can react to.

Those generated concepts can then become part of the broader engagement process. Participants can share their images in the gallery, while others can respond to and support the ideas they find compelling. For facilitators, our latest release also allows submissions to be pre-screened before they appear publicly, helping projects maintain appropriate moderation while keeping the experience seamless for participants.

We see this less as replacing the designer at the table and more as exploring how some of the creativity of a design workshop might be extended beyond the room.

AI-generated content in public engagement understandably raises new questions—about accuracy, expectations, moderation, and how generated images should influence a planning process. That is precisely why we are approaching AI  Sketch as a prototype.

We now want to put it in the hands of planners, designers, transportation professionals, and engagement practitioners. We want to learn where it is useful, where it isn’t, and what it would need to become a genuinely valuable part of the engagement toolkit.

This is an early step, but one we’re excited to explore: helping more people move from “Here’s what I’m imagining” to “Here’s what that idea could look like.”

‍