Trumpf + Compaile
High recognition rates > 99 %
Digitization: for paperless production
Enormous amount of time saved
“COMPAILE completely lived up to its promise of a development process based on open technology using open interfaces. They worked closely with our experts to complete the job quickly and professionally, with a total focus on solutions. The fact that COMPAILE offers a broad spectrum of expertise, from data labeling and AI to custom interfaces, is hugely beneficial!” Maximilian Veith, TRUMPF
The Trumpf Group is a company headquartered in Ditzingen, near Stuttgart. Trumpf is one of the world’s largest suppliers of machine tools. With more than 70 operational subsidiaries, the Trumpf Group is active in all major global markets. TRUMPF Werkzeugmaschinen GmbH + Co. KG produces and sells laser machines to many customers in the manufacturing sector; most customers are medium-sized companies.
For many of these customers, manufacturing is a fast-paced business; they often find that they need to change the components they are currently manufacturing based on the orders that come in – sometimes, they have to make those changes within a single day. They manufacture non-standard parts, some of which are only produced a few times a year, and new parts can be commissioned at any time. They still need to be able to recognize and categorize these parts. Consequently, one of the company’s most significant tasks is to correctly and efficiently categorize the parts on a consistent basis without the need for markings on the parts or for paper lists. Without this new option, it would be impossible to fully digitalize production to the degree required for a smart factory.
Tackling this central challenge requires a new solution that not only meets industrial requirements (versatility, durability, and flexibility at high levels of productivity), but also fulfills ethical standards (data privacy, operational reliability, etc). And ideally, it should do so in real time.
In a single training session, we used a neural network and complementary technologies to teach an artificial intelligence system to run a content-based comparison (rather than a visual one) on a large number of non-standard parts, and with BauIdent, it is now possible for the system to identify any non-standard parts without having “seen” them first, and to assign them to a blueprint. This technology can also assess the quality of individual manufactured parts by identifying differences between them and a sample part.
Our solutions are precisely targeted to solve our customers’ problems. Our development processes are rapid and based on open technologies; we work hand-in-hand with our customers and always keep their interests in mind.
more success stories
Entity recognition and media management based on a neural network in a web-based user interface at German government security agencies (domestic and international security)
Open, distributed, real-time-capable, and secure operating system for data-driven, AI-based applications in production.
Component recognition and quality assurance based on a content-driven similarity comparison at TRUMPF Werkzeugmaschinen GmbH & Co. KG