KISprachtec

AI-based language technology for NLG-based product information

Together with the Stuttgart-based language technology specialist AX Semantics, Reutlingen University is researching into how artificial intelligence can be used in the automated and scalable creation of texts from structured data across all processing steps.

In the joint project, scientists from Werk150, the teaching and research factory at Reutlingen University, are collaborating with AS Semantics to further develop the company’s ‘NLG Cloud’ software-as-a-service solution.

At Werk150, the project includes research on the Asset Administration Shell. This is a digital representation of tangible or intangible assets (e.g., processes or AI models in the field of Industry 4.0, which can be used as a structured database for text creation.

The aim of ‘KISprachtec’ is to optimise text creation on the basis of structured data, such as provided by the Asset Administration Shell. ‘Let’s take an electric bicycle, for example’, explains Professor Dr techn. Daniel Palm, who is supervising the project together with his team in Werk150. ‘The Asset Administration Shell creates a digital image of the bicycle which accompanies and updates it throughout the entire product life cycle. This starts with production and continues through to repair measures and recycling.’ For example, AI applications use the data from the digital bicycle to create assembly instructions or repair instructions.

During the development phase, the scientists always bear the transfer into the real world in mind. Werk150 is already working closely with small and medium-sized companies on numerous projects to support them in the area of digitisation and Industry 4.0. The infrastructure and work processes of the teaching and research centre are ideally suited to application-oriented projects such as ‘KISprachtec’.

This innovation exploits the ongoing merger of the real and digital worlds to tap into previously untapped potential from the end-to-end digitisation of the information chain between the product level and the advertising message.

The Invest BW programme from which the joint project is funded is an initiative of the Baden-Württemberg Ministry of Economic Affairs, Labour and Tourism. It aims to support companies in maintaining their innovative strength and initiating new research and development projects.

Visit the project page

By combining the Asset Administration Shell and artificial intelligence, we have the opportunity to replace static texts with dynamic and flexible content. ‘KISprachtec’ is thus responding to the growing demand for complex, personalised and optimised content in e-commerce and the manufacturing industry.
Professor Daniel Palm
Project manager