The bottleneck moved.
Filament winding has always been a process where geometry and manufacturing are tightly coupled. A pressure vessel may look simple from the outside, but the winding plan has to satisfy structural angle targets, pattern closure, dome turnarounds, finite tape width, polar buildup, and the machine axes that physically place the material on the mandrel.
In many teams, the slowest step is no longer creating a first path. The bottleneck is confidence. Engineers need to know whether a proposed winding plan is actually manufacturable, whether it builds the intended laminate, whether transitions remain continuous, and whether generated motion will make sense to the operator standing at the machine.
What gets missed when software stops at curves.
A centerline curve can hide several practical failures. The roving has width, so coverage and overlap are not the same as line spacing. The tape has thickness, so each layer changes the surface for the next layer. A helical layer may be mathematically valid, but the transition into a hoop layer can still create a discontinuity. A machine playback can follow the correct contact point but still orient the payout eye in the wrong direction.
These are not cosmetic issues. They affect whether the operator can trust the simulation and whether a machine owner is willing to try a generated program.
Why this matters now.
New pressure vessel programs, hydrogen storage work, low-cost winding systems, and retrofitted machines all push more responsibility onto software. Teams need fast iteration, but they also need traceability from design intent to machine motion. That combination is hard when planning, visualization, export, and customer feedback live in disconnected tools.
The better approach is to treat winding software as a verification environment, not just a path generator. The simulator should show the wound laminate, expose the weak assumptions, and produce an export package that a machine owner can dry-run before anyone considers real production.
Where AddWind is focusing first.
AddWind is being built around this design-to-machine confidence loop. The current pilot already focuses on realistic laminate rendering, pattern selection, transition diagnostics, project persistence, four-axis machine playback, tangent tape payout, and dry-run export packages. The remaining work is being tracked publicly so pilot users can see what is ready, what is being verified, and where their machine data can improve the product.
- Use standard projects to compare winding angles and pattern closure.
- Inspect completed laminate texture separately from live production playback.
- Save machine-specific project states before exporting controller code.
- Submit screenshots, controller formats, and machine constraints through the feedback board.
The practical buying question.
The question for a filament winding software buyer is not "can it draw a helix?" The better question is whether it can shorten the distance between an engineering idea and a machine-owner-approved dry run. That is where modern winding software has to earn trust.
The pilot program is designed around that proof: users bring real parts, real machine constraints, and real export requirements, while the product backlog is shaped around the highest-value verification gaps.