Munich Datageeks - September Edition

After a short break we’re back with our 50’th Meetup!

We are very happy to have such a long running meetup series, which wouldn’t be possible without our great speakers, sponsors and community! It's amazing to see so many people fascinated by data science just like we are. Thank You!

Our first time sponsor Oerlikon was so kind to offer bus transportation to and from their office.
Busses leave 18:15, 18:30 and 18:45 from Ostbahnhof (Friedenstraße 17a, back exit near Burger King) and go back there at 22:00, 22:15 and 22:30.

The lineup:

Ingo Thon - Challenges for using Machine Learning in Industrial Automation

Abstract:
Industrial automation utilizes control systems for handling different processes and machineries in industries. Most industrial systems are large scale distributed systems often interfacing with external systems like building automation. The first industrial automation started with the mechanization of the textile industry. The second revolution came Early 20th century when Henry Ford invented the assembly line and started mass production. In the third industrial revolution production went digital and CPUs took over control. In the current -- fourth -- industrial revolution industry is adopting novel technologies like IOT, cloud computing, 3d printing, robotics and most importantly artificial intelligence. Artificial intelligence and especially its sub-field Machine Learning can and will have a large contribution in this revolution. If we manage to overcome the challenges which arises when combining a largely academic field like ML and industrial automation a field whose paradigms have been developed over the course of the last century, like stability and reliability. One example is that communication subsystems are made for real time communication and not for large data transfers But the biggest challenge comes from the fact that the Machine Learning is not any more a single task but a tiny component of a large scale system

Bio:
Ingo Thon is the Principle key expert for Artificial Intelligence in the business unit factory automation. He is driving the adaptation of AI, where he has used Machine Learning for many different applications within SIEMENS. One of his largest successes is the SIEMENS inventor of the year award 2019. Before, working for SIEMENS he has worked in a startup as Software Developer integrating machine learning functionality in NoSQL databases and as Agile Coach.

Victor Ly - Using Big Data to Design Next-Generation Materials

Abstract:
Materials science is among the disciplines that can be coupled with and enhanced by big data – particularly in the design of new materials. Traditional alloy design is a laborious process involving gradual additions of alloying elements to a base material to study the impact of elements on microstructures and ultimately bulk material properties. Rapid Alloy Development (RAD) is a software platform through which large quantities of thermodynamic data are collected, mined, and visualized; this implementation of big data breaks the paradigm of material design. Features in the data can be analyzed and related to physical properties to predict actual performance in a variety of applications. Multiple properties are modeled simultaneously in thousands of distinct alloy chemistries to accelerate material design and testing. With the magnitude of materials spaces available, it is anticipated that big data will drive development of a new generation of high-performance materials.

Bio:
Victor is a programmer in San Diego, CA and has worked at Oerlikon Metco for 2 years. He is one of the main developers of the company’s Rapid Alloy Development (RAD) tool and is also deeply involved with the Smartfolio project – aimed towards analyzing the company product portfolio and improving the manufacturing efficiency of Oerlikon Metco’s production plant in Troy, MI. Victor graduated from the University of California, San Diego in 2017 with a background in materials science.