The Forum Junge Spitzenforschung awards prizes to the most innovative application ideas on the subject of Artifical Intelligence

Six young researchers receive awards

On 18 November 2020, the closing event and award ceremony for the winners of the sixth ideas competition “Forum Junge Spitzenforscher” of the Stiftung Industrieforschung in cooperation with Humboldt-Innovation took place as virutal event.

1st place went to PD Dr. Ludwig Schlemm (Charité – Universitätsmedizin Berlin) with her project: “AI-supported stroke detection”.

2nd place went to Kai Hoppmann-Baum and Jaap Pedersen (Technische Universität Berlin) for their project: “KOMASS”.

3rd place went to the project “PrivML” by Franziska Boenisch (Freie Universität Berlin).

The first three places will receive prize money of EUR 10,000 each.

A 4th place each were awarded: Volksbot: Gretchen (Heinrich Mellmann and Matthias Kubisch, Humboldt-Universität Berlin), Predictive Privacy: Innovative data protection for innovative AI (Dr. Rainer Mühlhoff, Technische Universität Berlin) and ZerOps – A Self-Healing AIOPs Platform (Dr. Florian Schmidt, Alexander Acker and SörenBecker, Technische Universität Berlin)

We would like to thank the top-class jury of experts as well as our Keynote Speaker.

Please visit our Youtube channel for videos about this year’s event.

Jury & Keynotespeaker

Jens Lambrecht

Jens Lambrecht studied mechanical engineering at the Technische Universität Dortmund. Already as a student, he worked in the field of simulation in industrial robotics and later received his doctorate from the Department of Industrial Automation Technology at the Technische Universität Berlin on the topic of human-robot interaction with gestures and augmented reality. He then spent half a year in Taiwan working in the field of image processing and then became team leader for cloud robotics at the Telekom Innovation Laboratories in Berlin. In 2016 he founded the company Gestalt Robotics and recently became an assistant professor at the Technische Universität Berlin. Since April 2018 he has been head of the Industry Grade Networks and Clouds department.

Trinh Le-Fiedler

Trinh Le-Fiedler can look back on over 15 years of professional experience: A lawyer with a Master of Laws from Harvard and a scholarship from the Fulbright and the German National Academic Foundation, she developed business models for deep-tech start-ups (as CFO & Head of Biz Dev @XAIN) before founding her deep-tech AI start-up Nomitri, led large sales teams (as Director @Wayfair), increased revenues (as Head of Buying/Sales @Wayfair), optimized operations (as Principal @BCG) and negotiated private equity deals as an attorney for KKR & Blackstone.

Rasmus Rothe

Rasmus Rothe is an internationally recognized AI and computer vision expert as well as co-founder and CTO of the Venture Studio Merantix, which promotes the transfer of AI research to business applications. Rasmus Rothe is also a founding member and board member of the AI Bundesverband e.V., where he provides impulses for the implementation and further development of the AI strategy of the federal government. In 2019 he was listed in the “30 under 30” list of Forbes magazine and in the “40 under 40” list of Capital magazine. At Merantix, Rasmus Rothe has been working since 2016 together with Adrian Locher (CEO) and a team of over 70 people to ensure that the research results of renowned scientists are put into practice. Under the umbrella of Merantix, through access to highly qualified talents, industrial partners and with the help of venture capital, startup teams can work on AI applications and thus create and benefit from a unique ecosystem. The current focus is on the healthcare, automotive, business intelligence and biotech sectors.

Jack Thoms

Jack Thoms is deputy head of the research group: Intelligent Analytics for massive Data and deputy spokesperson at the Berlin site of the German Research Center for Artificial Intelligence (DFKI). He is also head of the Digital Technologists Forum, a demonstration and networking platform for selected research projects and innovations in the field of digital technologies from Germany. Prior to his work at DFKI, he was head of department at Bundesdruckerei in Berlin, head of strategy and project development at the House of Logistics and Mobility (HOLM) in Frankfurt and lecturer and head of the Teradata Competence Center for Data Analytics at the European Business School (EBS) in Wiesbaden.

Roland Vollgraf

Roland Vollgraf, is Head of Zalando Research, a Machine Learning & AI research group within Zalando that is focussing on long-term strategic research questions around the Zalando platform. Roland joined Zalando in 2013 and has been playing a key role in the establishment of Zalando Research. Prior to that, he was Head of Research at GA Financial Solutions GmbH in Berlin, where he was responsible for the development of quantitative risk models and trading strategies in the capital market. Roland received the title of Doctor of Engineering Sciences in 2006 at the Technische Universität Berlin in the field of machine-learning and statistical signal processing.

Susan Wegner

Susan Wegner is responsible for Artificial Intelligence & Data Analytics at Lufthansa Industry solutions. She has more than 15 years of experience especially in the fields of Machine Learning, Artificial Intelligence and Platform/Software Design at Deutsche Telekom and other companies e.g. Robert Bosch, T-Systems. In addition, since 2013 she has held different positions within Motionlogic a subsidiary of Deutsche Telekom Group: Founder, CEO, chairperson and now member of the advisory board. Furthermore, since 2015, Susan has been a board member of the Bitkom Big Data Group, since November 2018 Member of the European Commission Expert Group on Business-to-Government (B2G) Data Sharing and judge/in the council of experts for e.g. The European Data Science & AI Awards or Bikom Big-Data&AI Summit. As a computer scientist and mathematician – studying in Berlin and North Carolina (USA) – she had the privilege to be an early adopter of Artificial Intelligence leading to her PhD.

Richard Socher

Richard Socher is the CEO of SuSea, a new startup that will bring trust, facts, kindness and science to the Internet to help people make complex decisions and will launch in 2021. Prior, he was the Chief Scientist and EVP at Salesforce and an adjunct professor at the Stanford Computer Science Department.
Before that, Richard was the CEO and founder of MetaMind, a startup acquired by Salesforce in April 2016. MetaMind’s deep learning AI platform analyzes, labels and makes predictions on image and text data so businesses can make smarter, faster and more accurate decisions.
Richard was awarded the Distinguished Application Paper Award at the International Conference on Machine Learning (ICML) 2011, the 2011 Yahoo! Key Scientific Challenges Award, a Microsoft Research PhD Fellowship in 2012, a 2013 “Magic Grant” from the Brown Institute for Media Innovation, the best Stanford CS PhD thesis award 2014 and the GigaOM Structure Award. He is currently a member of the World Economic Forum’s ‘Young Global Leaders’ Class of 2017 and was appointed to the Board of Directors for the Global Fund for Women.

Technische Universitaet Berlin, TUB,Fakultaet IV Elektrotechnik und Informatik, Institut fuer Softwaretechnik und Theoretische Informatik, Fachgebiet Datenbanksysteme und Informationsmanagement DIMA, Leiter Prof. Dr. rer. nat. Volker Markl

Volker Markl is a Full Professor and Chair of the Database Systems and Information Management (DIMA) Group at the Technische Universität Berlin (TU Berlin). At the German Research Center for Artificial Intelligence (DFKI), he is Chief Scientist and Head of the Intelligent Analytics for Massive Data Research Group. In addition, he is Director of the Berlin Institute for the Foundations of Learnig and Data (BIFOLD), a merger of the Berlin Big Data Center (BBDC) and the Berlin Center for Machine Learning (BZML). BIFOLD is one of Germany’s national Competence Centers for Artificial Intelligence and will further bolster ongoing collaborative research in scalable data management and Machine Learning. Dr. Markl is a database systems researcher conducting research at the intersection of of distributed systems, scalable data processing, text mining, computer networks, machine learning, and applications in healthcare, logistics, Industry 4.0, and information marketplaces. Earlier in his career, he was a Research Staff Member and Project Leader at the IBM Almaden Research Center in San Jose, California, USA and a Research Group Leader at FORWISS, the Bavarian Research Center for Knowledge-based Systems located in Munich, Germany. Volker Markl is a computer science graduate from Technische Universität München, where he earned his Diplom in 1995 with a thesis on on exception handling in programming languages. He earned his PhD in 1999 the area of multidimensional indexing under the supervision of Rudolf Bayer.


Franziska Boenisch

Franziska Boenisch

Freie Universität Berlin

The PrivML project will develop a framework that quantifies the loss of privacy in Machine Learning (ML) models and evaluates the privacy risk of individual data points. This will enable the legally secure and privacy preserving application of these methods. For although ML models are increasingly used for predictions on private data in sensitive applications, e.g. in the medical diagnosis context, this impact assessment has so far only been rudimentarily developed. Current research shows that it is possible to recover the training data from the models or at least to identify whether individual data points are contained in the training data set. Therefore, ML methods are currently being increasingly researched, which are supposed to ensure the privacy of the data in the models, e.g. by means of targeted noise. The problem that the effectiveness of these methods is currently not uniformly quantifiable is to be solved with the developed framework.

Kai Hoppmann-Baum &
Jaap Pedersen

Technische Universität Berlin

In order to achieve the Paris climate goals, decarbonization of all sectors is inevitable. In particular, the increasingly decentralized and volatile generation of energy by wind and solar power plants poses challenges for grid operators. Power-to-gas technologies play an important role in this context. The conversion of electricity into flexible, regenerative and storable energy carriers, e.g. hydrogen, serves to store surplus energy, use it as fuel or make it available as a basic product, e.g. in the chemical industry. However, the integration of these energy carriers into the existing system poses new and complex problems, especially for gas network operators. Our goal is to develop a decision support system that uses a combination of intelligent optimization algorithms to reliably provide safe and economical recommendations for the operation of the gas transport networks of the future, including new gas components.

Heinrich Mellmann &
Matthias Kubisch

Humboldt-Universität zu Berlin

Based on biological systems, research on embodied artificial intelligence (Embodied AI) is gaining increasing significance. Biological systems are inherently decentralized with individual parts of the body having their own intelligence. Intelligent behavior only then arises from the interaction of individual components and the interaction with the environment. Therefore, the body is an indispensable part in the research of intelligence. Of particular interest are humanoid robots, which are still difficult to access, costly and often have to be developed by scientists themselves, thus limiting the circle of users. In this project we develop a humanoid robot “Gretchen” for research and education. The focus is on the openness of hardware and software, as well as accessibility for researchers from other disciplines. Prototypes of the robot exist and are already used in courses at Humboldt-Universität zu Berlin

Rainer Mühlhoff

Dr. Rainer Mühlhoff

Technische Universität Berlin

Artificial intelligence not only has great potential, but also new challenges for data protection: Sensitive information about people – e.g. diseases, sexual orientation, creditworthiness – can be derived from easily accessible data such as tracking data, Facebook likes, browser history, etc.
The Predictive Privacy Project is developing a new concept of privacy protection to fill gaps in existing data protection regulation: The “predictive privacy” of an individual includes information that can be predicted about him or her through machine learning. Predictive privacy is violated when sensitive information about a person is derived without his or her knowledge and against his or her will and used, for example, for automated decisions such as determining insurance premiums or selecting job applications. In addition to the philosophical foundations of this approach, the project is developing a “Predictive Privacy” certificate for companies and research projects that want to guarantee their users the protection of their predictive privacy, and is interlocking this with new concepts for school teaching to promote public education about predictive analytics.

Ludwig Schlemm

PD Dr. Ludwig Schlemm

Charité - Universitätsmedizin Berlin

Acute strokes are one of the most common causes of permanent disability, leading to a loss of independence in everyday life and high expenditure in the health and social security system for those affected.
For a few years now, a complete cure has been achieved in many cases, even for severe strokes. However, it is essential for this to begin treatment as soon as possible after the symptoms begin.
In about 20% of all strokes, the symptoms occur unnoticed while the patient is asleep, which makes it impossible for this group of patients to start effective acute therapy early.
In our project we will develop an innovative solution for the real-time detection of severe strokes. The solution is based on non-invasive real-time monitoring of motor function, AI-supported detection of stroke-typical changes in the recordings, and the resulting early alerting of the emergency services.

Dr. Florian Schmidt, Alexander Acker & Sören Becker

Technische Universität Berlin

The constantly increasing system complexity and importance of reliable IT systems implies high demands on administrators. Artificial Intelligence for IT-Operations (AIOps) describes the process of maintaining and operating large and complex IT infrastructures using AI-based methods. This includes the automatic, early detection of anomalies and their causes, their correction and optimization, and the fully automatic execution of self-stabilizing activities. For this purpose we have developed the platform ZerOps as a research prototype, which uses AI methods to support administrators in the entire process chain: ZerOps includes ML models for anomaly detection, classification and selection of stabilizing solutions, as well as a decentralized execution platform for analysis steps to apply models and aggregate results in complex IT infrastructures.