LOGISTICS
Helmut Prieschenk, Managing Director of Witron, and Franziskos Kyriakopoulos, founder of 7LYTIX, discussed the potential for AI in logistics and demand forecasting for food retailers. Both agreed that AI offers significant optimization opportunities for streamlining processes in distribution centers and supply chains. The authors stressed, however, that high-quality data and individual experiences, as well as consumer requirements, are crucial for effective data modeling.
Taking note of the prevalence of AI hype, Pieschenk joked, “And then everyone was an AI influencer overnight.” In his view, AI tools should be evaluated for their ability to address specific requirements for customers and developers in a balanced manner.
Likewise, Kyriakopoulos cautioned against the spread of half-truths and stressed the practical applications of LLMs, especially in the processing of orders, sales, and customer communications. In areas like demand forecasting where some companies may have difficulty assessing the impact of AI models, he acknowledged the importance of transparency and understanding the added value of AI models.
While Prieschenk acknowledged the use of LLMs in Witron’s work, he stressed the importance of practical solutions over adopting new tools for the sake of it. His emphasis was on implementing stable solutions that optimize logistics processes in distribution centers and supply chains, while integrating well with existing ones.
In addition to their meticulous analysis of stored goods, sales, promotions, customer behavior, and location, Kyriakopoulos explained how they forecast demand accurately. While balancing accuracy with tangible value for customers and users, they incorporate machine learning methods progressively.
According to Prineschenk, AI models need to be aligned with logistics practical mechanics to ensure seamless integration. According to him, the supply chain should be viewed holistically, silos should be avoided, and explainability should be prioritized.
Human decision-making remains crucial, according to both experts, even though AI provides valuable information. They highlighted the importance of incorporating human experiences, gut feelings, and control over the process. Furthermore, they recognized the importance of delivering real value on a daily basis through stable models.
The future development of AI models for warehouses will build upon the existing achievements in revolutionizing logistics, according to Prieschenk.
