Generative AI for Data Insights on SAP BTP

Introduction

Generative AI is a branch of artificial intelligence focused on creating new data or content that resembles something it has been trained on. Unlike traditional AI systems that are designed for specific tasks, generative AI models learn patterns from a dataset and then generate new content based on those patterns. This content can take various forms, such as images, text, music, videos or even data analysis.

The terms cloud and cloud computing have been around for years. In simple terms, a cloud describes a network of systems that are operated elsewhere. Usually, a cloud is used to store data, e.g., Dropbox, Google Docs, whereby the structure and basic functionality of the underlying systems do not play a role. Cloud computing describes the use of the cloud infrastructure for storage and computing power to perform complex tasks such as building and using Generative AI models.

In this course, we will learn about the benefits of cloud computing and Generative AI and provide topics for practical development. As an example of how to run such software, in this course, we use the Business Technology Platform (BTP) for deployment and SAP Cloud Application Programming Model (CAP, https://cap.cloud.sap/docs/).

Contact

The course is offered by the Institute for Information Systems in cooperation with SAP.
DFKI: Alexander Berrang, Janaki Viswanathan
SAP: Dr. Christian Lander

Subjects

We will form 3 groups with max 4 students working on the same topic. Students can apply for one of the two topics (same complexity). We will try to assign topics according to your preferences. However, we cannot guarantee that you will get your preferred topic. The objective of each topic is to design and implement an analytics dashboard to be developed using the given data sets and the Generative AI PlugIn provided on the SAP Business Technology platform.

1. Individual performance recording at the digital manual workstation

Assembly processes in production must be trained to increase the overall performance, which can be checked through targeted measurements. In this context, we look at a manual workstation where cylinder heads are assembled. The assembly process comprises a number of different steps that must be followed precisely. We use various IoT elements to measure the activities of the worker. Your task is to use the event logs from several process runs to determine suitable parameters and visualize them in a dashboard that can be used to display the worker's performance.

2. Collective performance recording at the digital manual workstation

In this scenario, the manual workstation for cylinder head assembly represents the centerpiece of a training course. Several workers repeatedly carry out the process. Their activities are measured by various IoT elements. Your task is to create an analysis dashboard for the instructor based on the event logs of all the course participants and their process runs. The instructor should be able to view the performance of the entire course based on appropriate parameters. This includes the visualization and classification of aggregated data as well as the identification of interesting correlations in the sense of an associative analysis.

Important Dates

Basically, all dates take place on Fridays from 2 to 4 pm. To pass the seminar, attendance at all sessions is mandatory. Furthermore, a proposed Onboarding Guide has to be worked through successfully during the first 3 weeks to be admitted for the project. Also, during the group working phases, there will be a regular Q&A on the same Friday time slot, which the students have to attend.

Application

Due to the limited number of participants, you must apply for a place in this seminar. Selection will be based on your previous academic performance and other knowledge. Partizipation requirements include:

• Master's degree program in the Department of Business and Economics, in particular Business Information Systems and Digital Business Administration or Informatics
• Basic programming knowledge in Java and/or NodeJS

If you are interested, please send your application with your contact information, an excerpt of your previous academic achievements and a short note on another relevant knowledge to  alexander.berrang[at]dfki.de.