Our scrap went up.
Our floor opened up.

02 / Beyond the scrap bin

The work around the layout

That experience came back to me while reading Tim Heston’s “A Roadmap for Intelligent Nesting” in the October 2026 issue of The Fabricator.

Heston spoke with Lantek’s Victor Muñoz de la Parte and Anupam Chakraborty about current nesting capabilities and where AI could take them.

What caught my attention was how much of the article concerns the work around the layout.

How will the parts be unloaded?

What happens to the small pieces?

Will the blanks form correctly with the tooling available at the press brake?

Those questions reach beyond the scrap bin.

One example describes grouping 49 small steel parts into a mini-nest that suitable magnetic unloading equipment can handle together. That arrangement might sacrifice some material utilization while making unloading easier.

Whether it pays depends on the work saved—and any work added when the parts are separated and finished.

The article also describes sizing blanks using bend simulations based on the actual die widths at the press brake.

The nesting decision reaches into the next operation before the first part is cut.

My own question would go one step further: does the next operation receive what it needs to complete the job?

A sheet can be finished while an order is still waiting for one missing component.

That is why I want material utilization in the same discussion as handling, inventory, kit completion, and lead time.

Five percentage points of additional scrap deserve attention, especially on repeat work. Our experience does not make that loss unimportant. It shows why the scrap percentage alone could not tell us whether the change was worthwhile.

03 / The AI question

Where could AI help?

Heston’s interviewees describe a future in which AI could help predict processing time, machine performance, bottlenecks, and downstream capacity, giving optimization software better information for choosing a manufacturing route.

The article presents that as a direction of development. Many useful nesting capabilities already exist through established rules and optimization; they do not automatically require AI.

For me, the practical test would be whether an AI-assisted recommendation improves what happens on our floor.

Can it recover material yield without adding sorting work?

Can it keep unloading reliable?

Can it help deliver complete kits when the next operation needs them?

Can it protect the flow while improving the nest?

That is the challenge I would put to AI-assisted nesting.

I would want those five percentage points back.

I would also want to keep the 3,000 square feet—and bending in the morning.