Milestone / 2026.11.03

Autonomous Toolpath Generation

Testing new autonomous algorithms, fast-acting simulation and sensing, and real-time in-process metrology.

Toolpath generation remains an active area of research and development at ForgeNest. Alongside the toolpath tools that ship with the Forging Application today, we are currently testing:

--New autonomous algorithms for planning forging sequences
--Fast-acting simulation and sensing modules
--Real-time, in-process metrology

This work is in testing, not in production. It is aimed at a longer-term goal: a system that can evaluate a part during forging and adjust the remaining operations to stay on target.

Some of the goals we feel are achievable: the system will combine three AI subsystems working in concert:

Deformation Prediction Model -- A physics-informed neural network that will be trained on thousands of forging simulations and real-world sensor data. It will predict material flow and final geometry several steps ahead of the current operation.

Adaptive Toolpath Planner -- Given the predicted deformation state, this module will generate optimal press motions, die selections, and thermal management commands to converge on the target geometry.

In-Process Qualification Engine -- Using data from laser profilometry and embedded strain sensors, this module will continuously evaluate whether the evolving part shape remains within the qualification envelope.

Traditional forging relies on trial and error to develop process parameters for each new part. If these goals hold up in testing, they could reduce the number of test parts needed to bring a new geometry into production.

We will share results as testing progresses. Until then, Agility Forge customers have two production paths: validated part families through the Macro Module, and the Manual Toolpath Generator, where engineers plan the toolpath and the Agility Forge executes it automatically.

NextCapable System Online

All News & Media