JKU Is Accelerating Vehicle Software Development with Generative AI

Developing innovative features based on generative AI.

Hochreiter (right) with JKU and dSPACE employees; photo credit: dSPACE
Hochreiter (right) with JKU and dSPACE employees; photo credit: dSPACE

As part of a partnership between dSPACE (a leading solution provider to develop connected, autonomous and electrically powered vehicle) and the JKU's Institute for Machine Learning, dSPACE will be integrating generative AI technologies into its data-driven software solutions. Initial applications will be available by mid-2025 in “IVS”, dSPACE's software solution in sensor data management.

IVS - an open, cloud-based software platform – allows developers to analyze test data stored in the cloud, on-premise, or in globally distributed storage architectures. By means of natural language queries (prompts), the new generative AI-based capabilities can search automatically for items and actions across raw sensor data, significantly reducing the cost and time involved when developing new features.

Prof. Sepp Hochreiter, head of the LIT AI Lab and the Institute for Machine Learning at the JKU, remarked: "In the field of artificial intelligence, the partnership between dSPACE and the JKU's Institute for Machine Learning represents a symbiotic relationship between both the industrial and academic sectors. In re-focusing our research on industrial AI applications, we found that dSPACE's expertise in simulation, validation, and multimodal foundation models seamlessly complements our institute's strengths. The partnership is a unique opportunity to advance state-of-the-art AI research and address complex, real-world challenges. Our partnership will ensure that the JKU's research innovations will be seamlessly embedded into dSPACE's industry-leading tools, creating strong reciprocal advantages, especially as we continue to be at the forefront of developing generative AI technologies in Europe."

Tino Schulze, Vice-President of Automated Driving & Software Solutions at dSPACE, opens an external URL in a new window, added: "dSPACE’s partnership with the JKU is a unique solution to search for both identified and unidentified objects, and driving scenarios across our customers' vast data stores. The innovative approach will help our customers locate the right data from petabytes of unsorted data stores, as well as test and validate autonomous driving features by supporting generative AI technology. Our partnership will give engineers all over the world an opportunity to boost efficiency in developing with the support of generative AI technology."