Process-first metals research meets Data Science: Inside LKR at AIT’s pragmatic path from sensors to deployment

Session context and our editorial lens

In “Matthias Hartmann, Senior Research Engineer bei LKR Ranshofen / AIT,” Speaker Matthias Hartmann from the AIT Austrian Institute of Technology GmbH offered a clear window into a research environment where metal and materials science converge with data-driven practice. From the DevJobs.at editorial vantage point, what stood out was how firmly LKR links classical process expertise (casting, forming, additive manufacturing) with modern data capabilities (data engineering, data science, machine learning)—and how collaboration, communication, and a structured hiring process make that integration real.

With that line, Hartmann describes a culture grounded in real-world manufacturing processes while steadily integrating sensors, data pipelines, and flexible software deployments—up to and including Docker. The aim is not to do data science for its own sake, but to better understand and stabilize processes and channel insights back into production and simulation.

Who and what is LKR at AIT?

LKR (the Light Metals Competence Center in Ranshofen) is, as Hartmann stresses, interdisciplinary and anchored in metals and materials research. Teams are primarily composed of metallurgists, physicists, and mechanical engineers, with four teams in total:

• Three process groups focused on casting, forming, and additive manufacturing.
• One simulation team, chaired by Hartmann.

Across all teams, the need for data processing is growing: data engineering, data science, and adjacent skills are being deliberately built out. The rationale is pragmatic and process-centric: research increasingly involves instrumentation, data acquisition, and analysis—and those data should deliver tangible value by translating sensor signals into insights about process behavior, quality, or corresponding simulation models.

The organization has been around for roughly 25 years. It’s not a software house per se, but—in Hartmann’s words—“process people,” who orient themselves by real manufacturing activities. Simulation is deliberately aligned to processes rather than treated as a detached parallel world. That alignment is the intellectual draw: simulations that are grounded in real data, and processes that are tuned by simulations and data analysis.

Engineering culture: process proximity as a guiding principle

If you join LKR, you enter a culture with a clear spine: the team defines itself through processes and their improvement. Digitalization isn’t a “project,” it’s a trajectory that permeates day-to-day work. Priorities come through unambiguously:

• Process proximity: Every data-driven activity ties back to real manufacturing steps—casting, forming, additive manufacturing—and their practical challenges.
• Models with traction: Simulations stay tethered to process data and operational objectives.
• Practicality: Solutions have to run at the customer, on heterogeneous systems, and travel well across environments.

That culture guards against technology for technology’s sake. Machine learning, neural networks, cloud services—they’re means, not ends. They’re used where they extract reliable, actionable learnings from process data and hold up in plant environments.

The data value chain: from sensors to science

Hartmann outlines a data value chain that many modern manufacturing settings aspire to—at LKR, it’s uncompromisingly process-centric:

1. Sensoring and instrumentation: Processes are equipped with diverse sensors. Some of this happens in collaboration with sensor manufacturers. The team works “very close to the state of the art.” For candidates, that means exposure to new sensor tech and real process data.
2. Evaluation: Collected data are prepared and analyzed. Hartmann cites work with the company Eboard as part of this chain.
3. Data engineering: Data are filtered, cleansed, and quality-assessed. The goal is transferable, robust pipelines—not one-off analyses.
4. Data science: On that foundation, models and hypotheses are developed. The aim is to “get more out of the data”—concrete insights about the process or the corresponding simulation.

This shows LKR’s end-to-end perspective: from physical sensing to models that feed back into process and simulation. For data engineers and data scientists, this creates a field of play that doesn’t end in a dashboard; it translates into actual process improvements.

Technology choices: local, cloud, and containers—always with deployment in mind

Technologically, Hartmann emphasizes flexibility. LKR works with either local databases or cloud-based solutions depending on the use case—with Microsoft Azure explicitly named as an example. The platform choice follows the goal of making data available where needed and maintaining the ability to run machine learning and neural networks on the relevant platforms.

What matters is transfer to the customer environment. Results shouldn’t end at the LKR workstation. Accordingly, the team uses Docker to deploy solutions “independently,” avoiding environment lock-in. This is critical from an engineering perspective: containerization prevents a prototype that runs on a Windows development machine from failing at the customer, for example on an embedded Linux system.

That sentence gets to the heart of the culture: architecture decisions are made with the end goal in mind—shipping something that works in real plant settings. If robust engineering matters to you, that mindset will resonate.

Interdisciplinarity requires communication—truly

Hartmann highlights that the team is international and often works in English. In an organization that brings together metallurgy, physics, mechanical engineering, and data disciplines, communication isn’t a side topic; it’s a prerequisite for getting anything done. That’s why communication sits alongside technical excellence in their hiring criteria.

For candidates, the message is clear: strong technical skills are necessary but not sufficient. You need to explain ideas and results to colleagues in other fields, ask questions, defend hypotheses, and create clarity in meetings. That expectation aligns with the process focus: data and simulation work must be coupled back to realities on the shop floor—and communication is how that coupling happens.

Hiring: a clear process with grounded expectations

The hiring process is structured in a way that creates transparency and reduces friction:

1. Application enters AIT recruiting: Materials are screened in the recruiting department. There may be an initial phone call to clarify questions and assess relevance.
2. Forward to team leads: Suitable applications go to the technical leads.
3. At least two interviews: Where possible, one interview happens on-site in Ranshofen. Candidates can see the halls, the equipment, the environment.
4. Decision and probation year: Upon hiring, there is a classic probation year—clear, transparent, standard practice.

The on-site element is more than a courtesy. It aligns with LKR’s culture: processes, sensors, equipment—all of that should be seen if you are going to build data-driven solutions for it. This isn’t a lab that exists only in slides.

Onboarding: a checklist and a go-to buddy

Onboarding is intentionally structured. A checklist helps cover organizational steps and interactions with staff units up front. Each new team member also gets a designated buddy—an “Einführungspartner:in”—as a first point of contact for all questions, from where the fruit is kept to more specific topics.

These details shape culture. They signal that LKR invests in orientation and support—especially important in an international, interdisciplinary setup. From a DevJobs.at perspective, this is strong evidence of pragmatic people enablement: the organization creates structures that make arriving easier and accelerate learning.

What this means for tech talent

From Hartmann’s remarks, the takeaways for data engineers, data scientists, simulation experts, and software engineers are concrete:

• Work at the equipment level: sensors, real processes, real data. Not a theoretical construct—hands-on technology.
• End-to-end responsibility: from sensor to pipeline to model—and back into the process.
• Deployment matters: containerized solutions that must run at the customer, on heterogeneous hardware.
• Collaborative environment: interdisciplinary, international, often English—communication is a core skill.
• Structured frame: clear hiring process, onboarding checklist, buddy system, and a transparent probation year.

Together, these points make LKR attractive to candidates who value robust, practice-oriented work—people as comfortable in the halls as at the notebook.

“Process people” in a digital world: why that works

Hartmann’s self-description—“we are process people”—is memorable. It suggests groundedness, which in digitalization is an advantage. Data work is only valuable if it measurably improves processes. At LKR, that implies:

• Sensors over gut feel: decisions grounded in real measurements.
• Simulation with feedback: models are coupled with process data.
• Platform neutrality: solutions are packaged to run reliably on target systems—if necessary on embedded Linux rather than a comfy Windows dev box.

This discipline gives projects substance. For engineering profiles that care about quality, reproducibility, and transfer, it’s an ideal setting.

Working across disciplines: what good communication means in practice

When Hartmann emphasizes communication, it’s not about showmanship. Practically, it involves:

• Translating across domains: metallurgy, physics, mechanical engineering, data—each has its own jargon.
• A shared problem frame: which process question are we trying to answer with data and simulation?
• Expectation management: what can models deliver, and what do they require in terms of data quality?
• Clean interfaces: from sensor to database to notebook—who hands off what, in what form, at what quality bar?

That is where data work succeeds or fails. LKR appears to have internalized this—and so communication sits next to technical excellence in their expectations.

Why now? The momentum of sensors, cloud, and ML

Recent years have created a clear pattern: processes are instrumented, data becomes a resource, platforms like Azure are accessible, ML methods are capable. LKR is using that momentum—tempered by the groundedness of “process people.” The result is a setup that employs modern tools without being driven by them. For tech talent, that means a context combining technical depth with tangible outcomes.

That statement reflects a productive curiosity: data are not an end in themselves. They should answer questions—about the process, the simulation, and quality.

What we’re taking away from the session

• LKR is interdisciplinary and process-focused—three process groups and a simulation team.
• Digitalization is a core feature: sensorization, cooperation with manufacturers, end-to-end data chains.
• Data engineering and data science are being built out—with an emphasis on filtering, cleaning, quality, and robust insights.
• Technology applied pragmatically: local databases or cloud (e.g., Microsoft Azure), ML/neural networks, containerization with Docker.
• Cultural markers: international, often English, communication as a must-have.
• Hiring and onboarding are structured, transparent, and hands-on—including an on-site view in Ranshofen, a checklist, and a buddy.

Closing: a setting for people who want impact in the real world

“Matthias Hartmann, Senior Research Engineer bei LKR Ranshofen / AIT” showcased how the AIT Austrian Institute of Technology GmbH, through LKR, bridges classical manufacturing expertise and modern data work. If you care about casting, forming, additive manufacturing, and simulation—and you want to think end-to-end across data chains—this is a place where technology doesn’t stall at a prototype; it ships to customers, into equipment, into reality.

For tech talent, that translates into real processes, real data, real deployments—and a team that treats communication as a core competence. It’s a home for people who want to be “process people,” practicing digitalization in a way that sticks.