Overview

AI in production

Stumbling blocks, tips & tricks

The seminar on the use of artificial intelligence (AI) in production offers a sound introduction and practice-orientated guidance for the implementation of AI solutions in your company. The aim is to provide participants with the skills to independently identify use cases, support implementation and overcome typical challenges.

In a hands-on session, participants will learn first-hand about the challenges of modern data analysis. They will learn what to look out for in order to increase the chances of success of initiatives from the outset. To this end, we will endeavour to develop specific use cases together with the participants and identify key points for their success.

Seminar objectives:
  1. accompany implementation: You will gain the ability to accompany AI projects from planning to implementation and to mediate between engineering and computer science in an interdisciplinary manner.
  2. accompany implementation: You will gain the ability to accompany AI projects from planning to implementation and to mediate between engineering and computer science in an interdisciplinary manner.
  3. Understanding data types: You will gain an overview of the most important types of data and their special features in the context of AI applications.
  4. Select models: You will learn to select and customise suitable AI models for different use cases.
  5. Avoid stumbling blocks: You will learn how to recognise and avoid typical mistakes and challenges at an early stage in order to increase the chances of success of AI initiatives.
Seminarinhalte:
Tag 1: Weswegen alle gekommen sind. Modelle und Anwendungsfälle.

Der erste Tag ist ein „Deep Dive“ in den Werkzeugkasten der KI. Anhand konkreter Anwendungsbeispiele aus dem Produktionsumfeld schauen wir uns die wichtigsten Algorithmen des Maschinellen Lernens an. Ziel ist ein grundlegendes Verständnis relevanter Begriffe und Methoden zu Künstlicher Intelligenz, Maschinellem Lernen, Neuronalen Netzen, Data Science, Deep Learning & Co. Der Tag ist anhand der unterschiedlichen Arten von Daten strukturiert. 

 
Tag 2: Was meist übersehen wird. Das Entscheidende

KI wird häufig auf das Modell reduziert. Dabei steckt der weitaus größere Teil des Aufwands und damit auch des Risikos, dass das Projekt scheitert, in den vermeintlichen Hilfsdiensten: Daten und Deployment.

To the Agenda

Preparation:

Before the seminar, you will receive an online questionnaire to record the participants' prior knowledge, expectations and goals. This enables the seminar to be customised to the needs and interests of the participants.

Summary:

After completing the seminar, you will be able to independently identify and evaluate AI use cases in your company, support their implementation and recognise and overcome challenges at an early stage. You will be able to mediate between engineering and computer science in an interdisciplinary manner and know what is important when analysing data and selecting models in order to implement successful AI initiatives. 

The speaker

Dr. Max Schwenzer

Production Data Scientist at Voith Turbo

 What does AI look like in production?
It democratises computer vision. No need for expert programming for every use case. Simply scale visual inspection based on data, using the same standard model for all. With the help of the dynamic, open-source AI community, we can improve quality on our assembly lines.