Aibuild's FETS Cuts AM Simulation Time from Days to Seconds

The Problem: Simulation Bottlenecks Stall Large-Format AM Adoption

Additive manufacturing engineers have long faced a brutal trade-off. Simulation catches defects before they cost thousands in failed builds, but traditional finite element analysis for thermomechanical behaviour takes days to run. For large-format metal and polymer parts, that delay is unacceptable. A failed DED build can waste tens of kilograms of metal powder, dozens of hours of machine time, and the salary of an engineer who cannot get that day back.

Aibuild, the London-based AM workflow software company, claims to have solved this with FETS for Additive Manufacturing. The cloud-based Finite Element Thermomechanical Simulation tool simulates stress, distortion, thermal effects, and thermomechanical effects across DED, friction stir, and Material Extrusion processes. It handles both metals and polymers, predicting residual stress, warping, cracking, sagging, interlayer bonding, and adhesion.

The company says FETS runs roughly 10,000 times faster than conventional alternatives. I am sceptical of round marketing numbers, but the underlying claim is worth examining: if simulation drops from three days to seconds, the economics of large-format AM change completely. Validation: NIAR Puts FETS to the Test

Aibuild took FETS to the National Institute for Aviation Research at Wichita State University for independent validation. NIAR is not a soft touch; it serves aerospace manufacturers who cannot afford to get thermal process control wrong.

> "Thermal control has been one of the biggest challenges holding back metal additive at industrial scale," said Jeswin J. Chankaramangalam, NIAR Program Director. "For the aerospace manufacturers we work with, this represents a validated path forward: they can adopt large-format metal AM with the thermal process control they need."

NIAR's endorsement matters because aerospace does not adopt tools on vendor promises alone. The institute's testing focused on whether FETS accurately predicts thermal behaviour across the full build strategy, not just isolated process parameters. What FETS Actually Does

The tool integrates with existing CAD workflows and runs in the cloud. Engineers upload a part, define the build path and material, and get a thermomechanical prediction in seconds rather than days. This matters most for large-format parts where thermal gradients are severe and distortion is the primary failure mode.

Aibuild CSO Guy Brown was direct about why the company built it:

> "We kept hearing the same thing from engineers. I know simulation is the right thing to do, but I just cannot wait three days for an answer. A failed build is often thousands of dollars in material, machine time and energy, and hours of someone's time you cannot get back."

The speedup comes from algorithmic optimisation specific to AM thermal physics, not simply throwing more compute at the problem. Aibuild is also layering AI enhancements on top of this foundation, which NIAR noted will make the platform faster still as those capabilities mature. Who Benefits

System integrators are the immediate winners. Robotics integrators building custom DED, cement, and large-format polymer systems need rapid simulation to iterate on machine designs. Their customers, often companies entirely new to 3D printing, lack the expertise to troubleshoot thermal issues empirically.

Consider a single-use part: a custom tool to transport a large turbine on a truck. It prints once. There is no production run to amortise setup costs across. Getting it right first time is everything. Accessible simulation for these one-off, large-format parts removes a significant barrier to adoption. The Bigger Picture

FETS arrives as Aibuild expands beyond pure simulation into Aibuild OS, an AI-driven workflow simplification platform. The strategy is clear: own the entire AM preparation stack, from simulation to toolpath generation to machine control.

Whether the 10,000x claim holds under every real-world condition remains to be seen. But the direction is correct. Large-format AM will not scale industrially until engineers can predict outcomes before committing material and machine time. Aibuild is betting that speed of simulation, not just accuracy, is the missing piece.

M4S TAKE

My take: AI claims need scrutiny. The useful implementations reduce cycle time or defect rates in measurable ways. Vague promises about 'optimization' without specific metrics are usually marketing.

Simon McLoughlin