Project – AI4MultiDirectAM: Printing with Its Own Judgment
Adaptive multi-directional 3D printing powered by artificial intelligence: a new Czech-German project AI4MultiDirectAM, for which Entry Engineering Member of MATADOR Group has received a prestigious award.
The production of molds for composite parts has one thing in common: it is slow, expensive and difficult to modify during the process. Traditionally, molds are cut from blocks, milled and then fine-tuned. The entire procedure takes weeks, and the room for error is significant.
At Entry Engineering, we deal with this problem regularly – we manufacture molds for body panels, technical covers and other composite components. And this created the need to find a way to speed up production, reduce costs and waste, and at the same time enable quick geometry adjustments.
Together with the Technical University of Liberec, the Fraunhofer Institute IWU and other partners, we are therefore launching the AI4MultiDirectAM project, whose goal is simple: to print molds in a way that makes industrial sense – quickly, with high quality and with minimal human intervention.
How the technology works
The system is built around a large-format 3D printer mounted on a 6-axis robot arm, equipped with a screw extruder for applying plastic granulate. Unlike regular printers that build a part layer by layer vertically, this system performs multi-directional printing along curves, at various angles and according to load or geometry requirements.
This results in high demands on precision and continuous decision-making during printing. That is why the process is not controlled only by a fixed program – artificial intelligence plays a key role.
The AI continuously evaluates data from thermal cameras, LIDAR and other sensors monitoring the entire process. If it detects a problem with layer width, temperature or
deformation, it immediately proposes a correction that is applied in the very next layer. No waiting for an operator, no overheating, no repeated prints.
What makes this approach different
The project is not about “just another 3D printer.” The aim is to create a complete autonomous system that:
• operates without the need for direct supervision,
• collects and evaluates extensive process data,
• learns from previous mistakes,
• makes corrections before an error affects the final quality.
The extruder handles more than 20 kg of material per hour, printing from widely available plastic granulate and with almost no waste. Each print also receives its own digital twin – a complete data record of the entire process including all interventions and deviations.
Who is involved in the project
• Entry Engineering – building and testing the entire robotic system in a real production environment.
• TUL (CXI) – developing AI models and ensuring communication between sensors, the robot and the control system.
• Fraunhofer IWU – providing testing facilities and industrial 3D printing know-how.
• RTT Automation – contributing expertise in sensor technologies.
• Lakowa – validating the technology in the production of plastic parts for ambulances and rail vehicles.
The team connects industrial practice with research institutions and jointly explores the path toward autonomous, adaptive production control.
What this will bring
Entry Engineering currently produces approximately 20 molds per year. With the new technology, we aim to achieve more than double that amount – faster and at lower cost. Molds that used to take three weeks can now be printed in roughly five days.
Material waste will be reduced by up to 90%, and every part will have a complete data history. The technology will also not serve only internal projects – the facility will be open to external projects from design, architecture or the aerospace industry.
Printing that can adapt
The AI4MultiDirectAM project will run until 2027, but already the technology is taking shape that will shift 3D printing from simply executing instructions to active, adaptive control.
Not that the printer “thinks” – but a system that understands what it is doing and what could go wrong is a major step forward for automotive, aerospace and other fields where precision, time and repeatability matter.
For this project, Entry Engineering Member of MATADOR Group has received the prestigious German Business Award 2025 in the category of Czech-German cooperation. Read more.
