Amazemet adds AI automation to rePOWDER atomisation
Amazemet, based in Warsaw, Poland, has announced the integration of an advanced artificial intelligence model to fully automate its rePOWDER ultrasonic atomisation machine. The move is intended to address the need for constant, manual supervision of metallurgical processes, often required in industrial and academic R&D.
In particular, the company noted the suitability of this upgraded machine for use in high-throughput materials testing facilities.
“In most institutions, it is far easier to buy new equipment than to hire new technical staff,” stated Dr Łukasz Żrodowski, CEO of Amazemet, Adjunct Professor at Carnegie Mellon University. “Amazemet focused its efforts on limiting the time users must commit to atomising novel alloys. Our new AI process control delivers much more autonomy, allowing researchers to focus on discovery or supervise more devices and processes at the same time. “
The AI model utilises machine vision, processing a live video feed from a welding camera to analyse melt pool characteristics in real-time. It autonomously determines and adjusts torch position, power, and material feeding every 120 milliseconds. According to Amazemet, this continuous optimisation ensures high wettability of the atomised material on the sonotrode, a key element for an efficient process, resulting in the highest possible yield in the desired Particle Size Distribution (PSD).
The AI also controls the ultrasonics, overspray removal, atomisation atmosphere and gas flow.
The integration process
The integration of AI required a comprehensive machine overhaul, centred around a new Advanced Control Cabinet featuring an industrial-grade GPU and high-speed, industrial PLC. Its integration with industrial networks via API enables remote process control and monitoring.
The rePOWDER machine features a new, specially designed and optimised plasma source and connected feedstock feeders to track the quantity of processed material. The machine also features integrated gas recirculation with a passivation system for increased safety.
The AI-integrated rePOWDER was benchmarked using Ti-6Al-4V (Titanium Grade 5) wire, reportedly achieving production rates of up to 0.5 kg/h and a minimum of four hours of unattended processing. Amazemet aims to extend this autonomous operation to eight hours in the next year.
Future development
While Ti-6Al-4V is the benchmark, the company stated that it is already developing autonomous processes for other high-value materials, including NiTi for shape-memory applications and the C-103 (Nb) alloy for high-temperature applications.
Amazement is also developing new feeder systems for bar/rod, machining chips, and powder feedstocks, further expanding the machine’s autonomy and material flexibility.
