DEV Community

Cover image for OpenRoads vs Shapezo A Reproducible Workflow for Parametric Road Models and AI Context
Future Built AI
Future Built AI

Posted on

OpenRoads vs Shapezo A Reproducible Workflow for Parametric Road Models and AI Context

When I review a road model, I separate geometry from appearance and generation from evidence. OpenRoads Designer can produce an editable corridor from terrain, horizontal geometry, vertical geometry, templates, and engineering rules. Shapezo produces an AI-generated first scene after I select an area on a map. Both can support early work, but they need different provenance and validation records.

Start with a road model contract

I begin with a small manifest. It states the coordinate reference system, units, elevation datum, terrain date, station direction, design speed, road classification, design limits, and the intended use of the output. I also record which elements are surveyed, imported, designed, inferred, or temporary.

For OpenRoads, I include the terrain source, horizontal geometry, vertical geometry, template definition, template drop intervals, corridor name, feature definitions, design surface, drainage data, and quantity boundary. This makes a change reproducible. Another engineer can see which input produced a surface or a cross section instead of treating the corridor as an unexplained mesh.

Validate terrain before building the corridor

I inspect point elevations, breaklines, boundaries, triangle lengths, gaps, and areas where the terrain crosses an unlikely feature. I confirm that road edges, channels, and existing pavement are represented by the right three dimensional lines. I also check that the survey extends far enough beyond the design limits to support tie-ins and drainage decisions.
This is a dependency check. If the existing terrain is wrong, a profile can look reasonable while being based on the wrong ground. Side slopes, earthwork, drainage, and construction elevations will then inherit the same error.

Build the OpenRoads logic in layers

I create horizontal geometry first, then establish the vertical geometry. I review curve radii, transitions, design speed, sight distance, and connections to existing roads. I use the profile to control grades, vertical curves, bridges, tunnels, and intersection elevations.
Next, I apply a template that defines the cross section. The template may include lanes, shoulders, medians, curbs, sidewalks, ditches, pavement layers, and side slopes. Template drops handle zones where width, crossfall, or slope conditions change. The corridor then generates the design surfaces and feature lines along the route.
I test one change at a time. If I move the profile, I check corridor continuity, cross sections, drainage, superelevation, and earthwork. If I change a template parameter, I check the surface, side slope, and quantities. If I adjust an intersection, I inspect every approach and the low points around the curb returns.

Keep Shapezo as a scope experiment

With Shapezo, I save the selected map boundary, date, and question that caused the selection. The AI-generated scene can reveal whether the study should include a wider interchange, a railway, a river crossing, or a nearby neighborhood. It can also help a nontechnical group understand why a road project needs more context than one parcel.

I do not use the generated geometry as a corridor base. I do not calculate station-based quantities from it or extract construction points. When a feature becomes important, I replace it with current survey, GIS, owner records, or a controlled OpenRoads object. The Shapezo output remains a record of the initial scope question.

Version the linked objects

OpenRoads revisions should identify the changed object and the downstream checks it affects. A profile revision can change the corridor, drainage levels, side slopes, cross sections, earthwork, drawings, and construction data. I keep those outputs on the same design version and recheck them before issue.
Shapezo revisions are different. I may enlarge the map boundary or regenerate a context to test the scope. That is not a formal revision to the road geometry. Keeping the two histories separate prevents a visual draft from being confused with an approved design alternative.

Validation before handoff

Before sharing OpenRoads data, I review geometry rules, template ranges, superelevation transitions, surface continuity, drainage conflicts, stationing, feature definitions, coordinate settings, sheet updates, and extraction codes. I state the quantity rules used for earthwork. Before sharing Shapezo, I compare major roads, terrain, structures, and water features with current references and label the output context only.

Final view

I use Shapezo to define the place and the first question. I use OpenRoads to build a parameter-linked road model that can support review, quantities, drawings, and construction data. The reliable handoff is explicit: map frame, verified terrain, controlled geometry, tested corridor, and synchronized deliverables.

Top comments (0)