direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Inhalt des Dokuments

Es gibt keine deutsche Übersetzung dieser Webseite.

Student Team wins Atos IT Challenge 2019!

A group of three students from our Master Project: Distributed Systems are the winners of the 2019 Atos IT Challenge  in Paris. https://www.atositchallenge.net/

The topic of this year's challenge was to develop ideas and prototypes on how to use machine learning for sustainability. The task was to devise an innovative use case and build a prototype, leveraging Machine Learning, to address a sustainability topic. Our team showed how their idea can provide a benefit to the world as well as how the prototype could be further developed and taken to market.

Our team is made up by the three master students Ajay Kumar Mandapati, Kashika Mnocha, and Matthes Krull. They developed an application prototype called Farmero and were supported by Sasho Nedelkoski, Dr.Lauritz Thamsen, and Prof. Andrea Cominola.

As claimed by ATOS and the reviewers: "Farmero is an innovative and easy-to-use Application for small farms which delivers up-to-date and reliable disease-risk-assessments. To achieve this goal, Farmero utilizes state-of-the-art disease detection algorithms applied on infrared satellite images freely provided by the European Space Agency. Crop disease detection has been part of the latest research programs in German institutes and inspired us to our work. Farmero helps Farmers to reduce the amount of fertilizers and pesticides they really need to a minimum. Farmero gives them a precise evaluation of threads for every area in the field. Farmero not only creates powerful disease detection but also shows how modern machine learning techniques can easily be utilized by someone who really cares about nature.", www.atositchallenge.net/idea/farmero/

A video presentation of the idea is available on YouTube.

The teams has received a prize and the support by Atos executives to realize their ideas.



Zusatzinformationen / Extras


Schnellnavigation zur Seite über Nummerneingabe