What challenges await Austrian industry?

In 2023, industrial companies will continue to face a variety of challenges. Various challenges are currently Crises and cost drivers negative impact on the economy. The ongoing problems in supply chains and the rising costs of energy, raw materials and labour are currently the biggest obstacles to the success of companies.

The resulting reluctance to invest and innovate leads to a serious problem for artificial intelligence and its future use, such as Viacheslav Gromovfounder and CEO of AI provider AITAD, explains: "The increase in the cost of energy and raw materials is forcing the economy to implement price increases, but these are not universally accepted. As a result, companies are often stuck with these costs."

Gromov continues: "In addition, rising costs and inflation are leading to a reluctance to buy. Companies' business prospects have reached a historic low since the start of the energy crisis - the number of optimistic companies has never been so low." Criticism is also voiced in many places that the domestic extraction industry is partly to blame for the major challenges facing the use of AI.

But not every player in the domestic industry thinks this way: "I'm always amazed at the things that are criticised. I think that it is the state's responsibility to create suitable framework conditions. As an entrepreneur, I endeavour to make the best possible use of these. I don't see any need for the state to pursue a strategy for artificial intelligence; this should come from the private sector," says, for example Markus Loinig, CEO of Senzoro, a company that combines ultrasonic measurements and AI to reduce maintenance costs.

Sepp Hochreiter, Head of the Institute for Machine Learning at the University of Linz

According to a Deloitte survey are the biggest challenges that companies face on the road to using AI: Risk management, integration and implementation, high costs and data issues.

Keyword data: Data protection is often the main problem, but the inadequate quality of the data is also a recurring issue. Johanna Schmidt, who holds a doctorate in computer science and heads the research group at VRVis, believes that the use of trained AI to handle large amounts of data is a sensible approach. However, the step before that should not be neglected.

This first step is to harmonise the data to ensure accurate results. "It is not very effective to apply an AI to a faulty dataset," notes Schmidt. She emphasises that the approach of using a trained AI to handle large amounts of data is certainly promising and that there are experts at VRVis for this. Nevertheless, the step of data harmonisation is crucial in order to achieve precise results. Schmidt currently sees an interesting approach in the use of edge devices, which could provide a wealth of additional information in production processes. She hopes that this will increase awareness of the value of data science in the coming years.

What is needed, of course, are more experts. "Especially when it comes to combining unstructured data in the data lake and structured data in the data warehouse," says Manuel Hausjell, consultant at software developer TietoEvry. Last year, around a quarter of companies were looking for data scientists, AI researchers or transformation experts to fill their skills gaps.