On Monday, I wrote about the groundbreaking lecture that Thomas Pomorski and the engineers from Dyndrite delivered during the ICAM 2025 conference. Their completely new way of looking at additive manufacturing through the lens of software opened participants’ eyes to a new version of AM reality - one that still relies far too heavily on hardware.
And yet, it’s software that will ultimately allow 3D printing to enter true industrial-scale production.
That’s the central theme of the Recode AM series: reminding readers of an apparently obvious truth the 3D printing industry still tends to deny…
One of the topics Pomorski and Dyndrite discussed at the conference was software-defined delta qualification. It may sound highly technical, but the idea behind the phrase is both simple and revolutionary.
In traditional manufacturing, every change - a new powder, a new laser, a new geometry - requires a full requalification cycle. In practice, that means preparing a new set of samples, measuring the microstructure, comparing results, documenting the differences, and waiting months for approval.
The process consumes enormous resources, and for companies producing critical components, like Ursa Major, every modification formally resets the clock back to zero.
The new concept proposes a different approach: if the process itself is software-defined, and every parameter and outcome is recorded as data, we can clearly separate what has actually changed from what remains identical.
There’s no need to re-prove what has already been proven. It’s enough to analyze the difference - the delta - and automatically assess its impact on the final result.
This seemingly small shift has massive consequences. It opens the door to scaling AM in a logical, predictable, and economically sustainable way.
Instead of multiplying experiments, engineers can validate differences using well-defined mathematical rules. Every step is auditable, every result repeatable. In practice, that means six months of testing can shrink to a single week - and dozens of trials can become a few simulations and a series of automated measurements.
It’s also a new form of trust in manufacturing - one no longer based on intuition, but on data correlation. The system itself recognizes when a deviation is significant and when it’s merely statistical noise.
In that sense, delta qualification becomes not just a method but a philosophy - a new way of thinking about engineering in a world where the boundary between the physical and digital is disappearing.
Qualification automation in practice
Pomorski’s talk made waves not because it sounded visionary, but because Ursa Major showed the first real applications of this concept in the production of engine components.
Instead of building new qualification plans for every machine configuration, the company used code-defined process templates that could be adapted across different LPBF platforms.
For example, a nozzle element printed on an EOS M400 machine could be transferred in a short time to a Sapphire or NXG 600E - without having to requalify the entire process from scratch. The system analyzed differences in laser power distribution, scan dynamics, and powder parameters, comparing them with previously stored metadata. When those differences remained within tolerance, it automatically approved the process; when they exceeded thresholds, it triggered additional simulations or trial builds.
That programmability opens entirely new perspectives for AM. In a world where every machine, powder, and geometry can be defined in code, engineers can build automated qualification models that learn over time. As production data accumulates, algorithms detect patterns in machine behavior, create predictive models, and can forecast how a given change will affect microstructure or mechanical performance.
This isn’t artificial intelligence in the marketing sense - it’s process intelligence: the system’s ability to understand its own data and make logical decisions based on it. Dyndrite and Ursa Major demonstrate that it’s possible to move from a reactive model (“check after printing”) to a proactive one (“check before you print”).
Such a qualification process could eventually become the foundation for shared industrial standards.
In an era when every major organization works on different platforms and protocols, software-defined delta qualification has the potential to become a universal language - one that translates parameters between systems without losing meaning.
Pomorski emphasized that this isn’t a closed-door vision limited to Ursa Major’s labs. It’s an open architecture built on interoperable data formats and full transparency. In this sense, it mirrors the evolution of the aerospace industry in the 1980s, when digital data-exchange standards were introduced. Back then, those standards enabled global supply chains; today, they could enable global standardization of additive manufacturing.
The concept also solves another chronic AM problem - the lack of consistency between the world of simulation and the world of real-world production.
Until now, engineers treated simulations as a preliminary stage that always had to be “validated on the machine.” Thanks to software like LPBF Pro, simulation data, sensor readings, and production logs now flow into a shared repository. That means every build becomes an experiment feeding the next one.
Over time, this creates a self-regulating ecosystem: data from previous qualifications helps shorten the next ones. That’s the true value of the delta concept — the ability to accumulate knowledge. The goal isn’t just to make testing faster each time, but to reach a point where tests are no longer needed, because the system itself recognizes process stability.
In the long term, such automation opens the path toward real-time certification. If every layer, every laser trajectory, every power and temperature measurement is captured and analyzed continuously, we can talk about a dynamic model of trust. The certificate stops being a static document - it becomes a stream of data.
Dream of scaled additive manufacturing
Today, when nearly every AM company claims to be “ready to scale,” the difference between marketing and reality often comes down to one question: can your process qualify itself?
That’s the essence of software-defined delta qualification. It’s not automation for automation’s sake — it’s about building a system that knows what it must prove on its own.
Ultimately, that might be the real revolution the industry has been waiting for. Not a new laser, not a new powder, not a new machine — but a new definition of certainty.