FACE TO FACE
Processed metal sheets at INTECH . Laser-cut sheet metal parts for automotive production .
our own productivity on the shop floor ,” continued Zimer .
AI can also be applied to increase the quality of a machine build . In the final assembly stage of a TRUMPF laser machine , a microphone records the sound of the operational machine . The AI interprets the recorded sound and analyses it for any anomalies ( i . e . a vibration , a missing screw etc .) so that this can be corrected . TRUMPF has also applied AI in image recognition e . g . for hairpin laser welding in electromotors .
“ Another example is preventive maintenance . We can monitor the condition of laser machines at a customer ’ s site to check functionality and to proactively act in case of failures . We get very positive feedback from customers on our AI solutions , as it helps to raise their productivity , increase their yields and maintain machine uptimes ,” Zimer told the audience at INTECH .
Looking to the future
However , AI is also a disruptive technology that will engender changes to the way that
TRUMPF TruBend 3100 press brake at INTECH .
Sheet metal battery housings .
manufacturers , such as TRUMPF , operate and will affect business models .
“ The speed of capability and innovation of AI is so rapid that we must conduct our research and development at a much faster pace than previously . We must also apply a trial-and-error approach because AI is not 100 % precise . This means that we need to change the way that we develop [ products ]. If you get the basics right for a problem to solve with AI , it is shocking how rapidly you can scale ,” explained Stephan Mayer .
“ AI will certainly change the business model for TRUMPF . It is not just selling machines ; it is a question of not leaving the customer with process responsibility . We are now responsible for the result on the shop floor . AI ’ s speed of adoption is incredible . In the end , we are managing data . It is such a powerful concept that we see tremendous speed of application inside the company and out ,” added Hagen Zimer .
TRUMPF ’ s ambition is to sell customised smart factories ( machines , systems , robots etc .). These systems can only be tested digitally : a factory simulation on a digital twin was demonstrated at the INTECH exhibition . A single part can be run for a customer in this
virtual environment to identify issues , cycle times , obstacles , risks and bottlenecks . This would be impossible without AI algorithms and the computing power available today .
However , AI networks are not yet selflearning and cannot provide analytics . Identifying root cause and optimisation loops will be the next level for the technology .
“ If you want to reach the next level of efficiency , it is important to work with customers . We know our machines and the customer knows his processes and requirements . We will never reach the ‘ sweet spot ’ if we do not exchange data . There must be a certain level of trust . We invest millions of euros into cyber security . We jointly interpret the data with the customer and jointly condition the AI algorithm — it is a shared task ,” said Stephan Mayer .
Despite challenges such as economic downturns and the softening of the Chinese economy , TRUMPF ’ s Management Board has decided to speed up its AI efforts .
“ There is no time to lose . AI is a tool that enables us to act faster in a changing environment and improve our internal processes where it makes sense ,” concluded Hagen Zimer . n
www . trumpf . com
To watch the full AI debate during the INTECH exhibition , see www . youtube . com / watch ? v = R1GMVuMkhXY
To watch TRUMPF ’ s panel at INTECH on mastering current and future industrial challenges ( such as digitalisation , the shortage of skilled workers and energy costs ), see www . youtube . com / watch ? v = dsJYT5IwI1Q
34 | ismr . net | ISMR May 2024