information desk

Newsletter from September 2022

Topics: ICML 2022 | dida new member of KI-Park | Projects LaserSKI & KIOptiPack

Dear dida follower,

From July 17th - 23rd this year's International Conference on Machine Learning (ICML) took place in Baltimore, USA. Besides the worldwide dominating AI companies like Meta, Google or Microsoft, which present their latest research results, dida was also represented with a paper and a talk this year. Lorenz Richter (PhD), our CTO, presented latest findings on "Robust SDE-Based Variational Formulations for Solving Linear PDEs via Deep Learning", which can be found here. We are especially pleased that we, as a comparatively small company, were able to present our results on important international stages for the third time in a row. But not only Lorenz was on the road for dida at current conferences: also our senior ML scientists Frank Weilandt (PhD) and William Clemens (PhD) presented project insights at the Living Planet Symposium of the European Space Agency (ESA) as well as at the Applied Machine Learning Days (AMLD) in Lausanne. For a general status quo on current AI developments, feel free to check out our summary of the AI Index Reports 2022.

Also worth mentioning is that dida has been a member of KI-Park since last quarter, an association of representatives from business, research and politics with the mission to make Germany and the EU a globally leading innovation location for AI by 2030. Together with companies like Deloitte or Volkswagen, we are excited to drive AI developments in Europe and contribute our expertise and perspectives. The AI Park was founded in 2021 and is still looking for members - maybe this network is also interesting for your company?

Interesting project developments include the LaserSKI and KIOptiPack projects. LaserSKI, which we have been leading since April 2020, was successfully completed last month. The research project in cooperation with the Ferdinand Braun Institute as well as companies from laser production was about detecting defects in different stages of laser production. Details about the project can be found here. The three-year KIOptiPack project started this month. In a consortium of science and industry, the goal is to optimize product design and production of plastic packaging with regard to sustainability aspects - dida is responsible for the machine learning development.

So much for our highlights from Q2/Q3 2022 - it's nice that we can update you regularly about developments at dida and ML in general. If one of the topics has sparked your interest to talk about it in more detail afterwards, please feel free to contact us (or have a closer look at our two offers ML-Expert Talk and Tech Lunch).

Best regards and see you next quarter

Philipp Jackmuth & Lorenz Richter