BiGmax Workshop 2019 on Big-data-driven Materials Science

Scope
Materials science is entering an era where the growth of data from experiments and calculations is expanding beyond a level that is properly processable by established scientific methods. Dealing with this big data is not just a technical challenge, but much more: it offers great opportunities. Big-data analytics will revolutionize new material discovery and will make the successful search of structure-property relationships among multiple length and time scales possible. This workshop represented the annual meeting of the Max Planck Network on Big-data-driven Materials Science.

Structure and Special events
Representatives of all projects reported on progress of their research, in addition there was a poster session. Special events included two tutorials on Deep Learning and Compressed Sensing, respectively. There was also a special break out session with dierent discussion/question rounds without the PIs (during the PI meeting) which was received very well by the junior participants.

Important participants
As an internal workshop of the network, there were only two external invited speakers:
Prof. Anatole von Lilienfeld (U. Basel, CH) and Dr. Alpha Lee (U. Cambridge, UK). Both have presented impressive accounts of data-driven work on small-molecule chemistry using machine learning techniques, partially far a way from the well known concepts.

Scientic newcomers
One notable newcomer was present: Markus Kühbach, BiGmax scientic advisor for software engineering. Kühbach stressed the different roles he represented within the network, the help he could provide in terms of clean, well-built software. He also mentioned recent guidelines on FAIR data infrastructure - a timely initiative that is bound to take a larger role within BiGmax.

Summary
The workshop this year helped bring the experimentalists with the theoreticians closer together. We have seen a lot more interactions between members, as well as a common language and framework shaping up throughout the talks, suggesting more coherence in the network. Smaller discussion rounds focused around technical topics helped students and postdocs exchange more intensively. The invited talks were regarded very positively as inspiring research directions. Overall we have received positive feedback from various members of the network.