This led to collisions in the machine and to production stoppages. With the help of special sensors and a KI algorithm, Trumpf has now found a way to avoid these disruptions. December 2018 Calendar Marathi Trumpf also looks into machine noise as an indicator of potential disruptions and failures, as Klaus Bauer, head of basic technical research at Trumpf, explains.
December 2018 Calendar Marathi
Explained in his speech at the Handelsblatt Congress. Recently, there is a special smartphone app for it. In a first step, the machine sends the microphone-picked sound spectrum to the Azure Cloud, where it is analyzed using an AI application.In addition, various background discussions and talks were held at ministerial level (Germany, EU). An indicator of the activities of the countries / regions are, according to Harhoff, the (published) contributions to important AI contributions. As expected, the US leads by a considerable margin, followed by the EU, with Germany being the most active journalist in the EU (see picture gallery).
For example, vibrations can be identified that the ear can not recognize. From the cloud, the analysis results finally reach the app. The service technician can then encircle the problem and decide whether and which maintenance or repair measures have to be taken. KI needs to train the algorithms n data that are not big enough. And of course, computing power.
The latter does not lack in the cloud age, in many cases already. The situation could be better, according to Klaus Bauer, if more companies decided to give their (machine) data to the cloud. But the data quality leaves something to be desired. According to Dr. Bernd Heinrich, CDO for Smart Mobility Solutions at Bosch, estimates that around 80 percent of the data collected is simply worthless, which is why the reservoir of usable data for the training of AI engines continues to shrink.
An alternative could be training with simulated data. Trumpf manager Bauer has as his data source the digital twin – that is the virtual image of a machine or plant – in the sights. With this “digital twin” certain machine conditions can be simulated., The generated data can then be used for training the algorithms and neural networks. How “top” is Germany at KI? But where does “Germany” stand up in terms of artificial intelligence specifically in international?
Answers to this burning question were tried at the Handelsblatt Congress Dietmar Harhoff of the Max Planck Institute for Innovation and Competition. As a basis, he used the EFI Annual Report 2018, which reflects the state of things in autonomous systems and artificial intelligence in a global comparison. The investigation is based inter alia on technical discussions with representatives of established companies, start-ups, research institutions and public institutions in Germany and the USA (Silicon Valley).
Similar findings are provided by a survey conducted by the VDI (Association of German Engineers) among 900 members , So Germany’s position in the global AI scene is not really bad, but not really “great”. December 2018 Calendar Marathi According to 80 percent of respondents, top AI nations are the US, followed by China, with China setting the pace in its AI China 2025 plan. Germany comes in third with a vote of 30.4 percent.