Version 1

Talk (30 min): How Machine Learning is making harbours safer

Learnings from building an easy to use Machine Learning Platform

Join online:

The basics of machine learning are already known for a while now. However, only with the development of high-capacity computers can machine learning be used in the mass market. The fields of application are very broad, and currently a lot of research is conducted. Even though this new research promises improvements in almost any sector. Companies nowadays struggle applying it. This is mainly due to three reasons: the lack of trained talents, the high costs of development and complex deployment into existing enterprise systems. BSI wants to change this by creating a machine learning platform that gives a simple interface directly embedded in the business environment – with a deployment process as simple as one button. It can be used without being an expert in that area, however, still with the flexibility of manually programmed machine learning workflows. In this talk we would like to present to you our motivation, the phase of development and the result of this project. We will illustrate the capabilities of our software using a real life example. At the Harbour of Hamburg we are helping to face the existing and highly complex challenge of maintaining cranes.

The development of a machine learning platform which is available to a broad mass of users enables a fast distribution of those technological possibilities. With this platform BSI empowers companies to improve their operational processes. Furthermore, we support marketing agencies in their targeting or data scientists in creating their own machine learning models. During this talk we would like to present a few examples that show how diverse this platform can be applied and how technically mature the platform already is. A special focus will be laid on the software solution created for the Harbour of Hamburg which supports the maintenance of the cranes. Additionally, we will grant a worthwhile insight into our thoughts about the design and share our experience with the different ideas of approaches we had. We will share what went well and where we still need to improve.


Day: 2020-10-10
Start time: 13:00
Duration: 00:40
Room: Online / ETH (RED)
Track: Computer Science


Concurrent Events