Overview
We are developing a marine-specific CAD system that allows designers to explore families of related boat designs while maintaining constraints on geometry, physical properties, and manufacturability.
Boat design is entering a period where historical precedent is no longer enough. Electrification fundamentally changes the weight, size, cooling, ventilation, and maintenance requirements of a boat’s propulsion system, while also altering the trade-offs among power efficiency, displacement, range, stability, and use of deck space. Rethinking vessels to accommodate these changes requires more than incremental refinements of existing forms.
Polymorph is a system built specifically for computer-aided exploration of boat hull design. In a traditional CAD workflow, a designer develops a geometry, runs an engineering analysis, adjusts the design, and repeats the process. Depending on the complexity of the analysis, meaningful feedback on a design change can take hours or even days, limiting how much of the design space can be explored.
Polymorph takes a different approach. Rather than focusing on a single hull, designers define a family of related geometries and guide the computer in a search for designs with desirable properties. Simulation becomes part of the design process itself, helping identify shapes that meet specific goals for stability, drag, weight, volume, manufacturability, or other criteria.
We see the designer’s role less as producing a single optimized hull and more as defining the boundaries of a design space. Within those boundaries, the computer can simulate performance, explore alternatives, and provide feedback on how a design might change to better meet its objectives. The goal is not automation, but a tighter feedback loop between human judgment and computational exploration.
Manufacturing is part of that feedback loop. A design is only useful if it can be built, whether through sheet-based construction, moulded construction, or other fabrication methods. Rather than treating manufacturability as a downstream concern, Polymorph aims to keep geometry, performance, and buildability connected throughout the design process.
Traditional CAD kernels were developed decades ago for broad industrial use. They are powerful, but they were not conceived with tight integration between geometry and real-time physics in mind, nor with modern optimization techniques that rely on structured, differentiable geometry. Polymorph is our attempt to provide that foundation.
The emphasis is on interaction. The human defines intent and boundaries; the computer explores within them and reports back with physically meaningful guidance.
Polymorph is not intended to replace existing CAD software for general use. Writing a new geometry kernel is a long-term research effort, not a commercial shortcut. Its scope is deliberately narrow: marine applications in which electrification and new hull concepts require tighter integration among form, physics, and fabrication.