AI Rescued Our Production Planning. Now We Build More Machines Without Hiring, Says Jan Slavík, Bednar FMT

July 2, 2025

Bednar FMT is a Czech family-owned company that develops and manufactures advanced agricultural machinery—especially large, mounted equipment for soil processing, seeding, and fertilization. As the company shifted from cell-based to line production, it encountered new planning challenges that traditional ERP tools couldn’t solve. In this episode of the Adastra podcast, Jan Slavík, CIO at Bednar FMT, shares how advanced optimization and AI helped turn chaos into measurable results.

Read the podcast as an interview

Ivana Karhanová: What happens when you manufacture massive agricultural machines and switch from cell-based to line production? How can AI and optimization help? That’s what we’re discussing today with Jan Slavík, CIO at Bednar FMT. Welcome.

Jan Slavík: Hello, and thank you.

Ivana Karhanová: Let’s begin with your company. What does Bednar FMT do?

Jan Slavík: Bednar FMT is a Czech family business that manufactures agricultural machines and technologies. We don’t just produce equipment—we deliver complete technologies for soil cultivation. That’s what the “FT” stands for: Farm Technology. Our machines are designed to fit into a full agricultural workflow, not just standalone tools.

Ivana Karhanová: To clarify, you don’t make tractors—you make the large implements that go behind them?

Jan Slavík: Exactly. We make mounted tools for soil processing, seeding, fertilization, and mulching—but not plows. We specialize in no-till technology.

Ivana Karhanová: And some of your machines are really huge, right?

Jan Slavík: Yes, they can be. Our range goes from 3-meter-wide machines to 18-meter-wide ones. Larger machines bring more value, and there’s less competition in that segment. It’s also a big engineering challenge—because even the biggest machines must fold down to a legal size for road transport.

Ivana Karhanová: Let’s talk about the production side. What changed when you moved from a cell-based setup to a production line—and how did IT and optimization come into play?

Jan Slavík: At first, it didn’t seem like an IT topic. But everything in the factory is run by IT systems. In the old model, we had “nests”—all parts were brought to one place, and workers assembled the whole machine from start to finish. It worked, but it was hard to track progress and required highly skilled labor.

We were growing fast and wanted a more efficient setup. Inspired by the automotive industry, we switched to line production. Each workstation now handles a specific part of the process, and the product moves forward step by step. It allowed us to specialize tasks, improve tracking, and increase transparency. That’s when IT became critical—to manage logistics, part flows, and real-time data from the floor.

Ivana Karhanová: So, you had to make sure people and parts aligned with the line?

Jan Slavík: Exactly. Everything has to arrive at the right place, at the right time. We already used barcode scanners, but we needed the whole process to be time- and location-synchronized.

Ivana Karhanová: When did you realize something wasn’t working?

Jan Slavík: Right after implementation. Instead of increasing output, the performance dropped. That wasn’t the plan. We quickly realized we underestimated how deeply this would affect work organization and material flow.

In the old setup, a missing part didn’t stop production—the team could work around it. In the line setup, everything stopped. The line would block because one piece wasn’t ready, and suddenly nothing could move forward.

Ivana Karhanová: And planners couldn’t react fast enough?

Jan Slavík: We had three planners handling dozens of stations and hundreds of dependencies. A single change could take half a day to replan. But by then, something else had already changed. It became chaotic. Workers followed verbal instructions from supervisors instead of the plan, while logistics continued delivering based on the outdated schedule.

Ivana Karhanová: So, the line ran, but it was inefficient?

Jan Slavík: Exactly. We were producing, but it was chaotic and far from optimal. The inefficiency became unsustainable.

Ivana Karhanová: That’s when you realized you needed something more powerful?

Jan Slavík: Yes. We needed a planning tool that could recalculate options in real time and find better solutions than humans could. Our ERP system couldn’t handle that.

Ivana Karhanová: And that’s when you turned to Adastra and implemented OptiSuite?

Jan Slavík: Right. But before any optimization could happen, we had to get our data in order. We thought we had good data—but when we loaded it into the system, it crashed. We found inconsistencies, missing steps, and gaps. The reality on the shop floor didn’t match what was in the system.

Ivana Karhanová: So, people were improvising based on experience?

Jan Slavík: Exactly. Workers had internalized the correct steps—even if the system didn’t reflect them. But optimization needs precision. We had to define every station, all constraints, configurations, shift patterns—everything. It took about three months.

Ivana Karhanová: And once the data was clean?

Jan Slavík: Then the optimizer worked. At first, people didn’t trust the system. But once they saw the plans matched reality—and actually helped them—they started relying on it. Logistics knew exactly what to deliver and when. Tasks became clearer. And we stopped wasting time on manual replanning.

Ivana Karhanová: How often do you update the plan?

Jan Slavík: At first, just twice a day—morning and noon. Later, we tied it to production progress. As soon as a task was reported completed, the system could adjust downstream plans. That helped account for delays or quality issues much faster.

Ivana Karhanová: What was the outcome?

Jan Slavík: We significantly increased output with the same number of people. That’s the key result. When we looked at the numbers, we estimated an added value of up to €10 million in revenue growth over five years.

Ivana Karhanová: That’s a huge gain.

Jan Slavík: And it improved safety too. Before, materials were delivered too early and cluttered the workspace. Now they arrive just in time, reducing hazards and confusion.

Ivana Karhanová: What’s your vision for the future?

Jan Slavík: We want a fully digital and autonomous factory. Eventually, workers will come in, check a screen for their assignment, and autonomous vehicles will deliver their materials. We’re not there yet—but it’s where we’re heading.

Ivana Karhanová: Was it hard to get people on board?

Jan Slavík: The hardest part wasn’t technology—it was changing mindsets. We had to make it clear that production is driven by process and planning, not just frontline decisions. Once people saw it worked, they accepted it.

Ivana Karhanová: Why not use an off-the-shelf solution?

Jan Slavík: Most tools are designed for serial production. Our product portfolio is too diverse—some machines take 30 hours, others 500. Some are 3 meters wide, others 18. We needed something flexible. Off-the-shelf tools couldn’t handle that without major customization. OptiSuite gave us a powerful optimization core we could build around.

Ivana Karhanová: What’s your biggest takeaway from the whole journey?

Jan Slavík: Don’t underestimate the state of your data. Before you even start thinking about optimization, check your data—and check it again. And be ready to align people and processes around it.

Ivana Karhanová: And the hardest part?

Jan Slavík: Changing behavior. Teaching people to follow the process as planned—not what they’re used to. That was the toughest piece.

Ivana Karhanová: Thanks for the insights, Jan. Fascinating story.

Jan Slavík: Thank you for having me.