FORVIA's Zvonimir Zaja on using AI and machine learning to manage fluctuations in demand скачать в хорошем качестве

FORVIA's Zvonimir Zaja on using AI and machine learning to manage fluctuations in demand

As well as delivering sessions on AI and smart automation at ALSC Digital Strategies Europe, Zvonimir Zaja, head of global supply chain and logistics at FORVIA Interiors, joined Automotive Logistics on the Red Sofa to discuss the ways in which FORVIA has improved its planning processes through AI and machine learning. "We've been working in the environment of demand uncertainty to improve our decision-making within our planning process using AI simulation," explained Zaja. "So we have built a digital twin of our planning environment and this allowed us to run thousands of simulations in a few minutes to understand the impact of our decisions before we make a decision." He noted that this resulted in an 8% reduction in inventory at its pilot plant, while experiencing no issues with its customers, as well as no shortages with its suppliers. Scaling AI When scaling up the use of AI and digital twin technology, Zaja said that trust is key. "You have to build trust," he said, with people and with the community so that everyone understands the model, what it does and what data it needs. On this point about data, he added: "We also have to make sure that the data is delivered, it's available, it's accurate." He also said that it is important to understand how that system is integrated into a company's ERP landscape. Applications of AI in the supply chain "Our use case that we have been doing in the field of demand planning, demand arbitration, is showing that this [application of AI technology] is feasible, it's possible," Zaja said. Looking elsewhere Zaja identified inventory optimisation and predictive analysis as two key areas in which AI can be "very valuable". However, he noted that the revolutionary nature of AI technology can sometimes be overstated, with regards to its ability to surpass and replace human judgement. "I don't see that [happening] yet," Zaja stated. "And also [in terms of] a fully autonomous running supply chain, I think we are not there yet, so still a way to go." Zaja shared his perspective that the aspects of supply chain planning that still require the greatest level of human intervention include building models and making final decisions. He explained that AI tools can be incredibly useful to give recommended actions, but ultimately the decision on which action to take should still be made by a human. Upskilling for digitalisation Zaja explained that when it comes to upskilling along an AI and digitalisation roadmap, there is "a very clear direction towards data management [and] data mastery." He said that people will need to be able to explain the models, and understand what is happening inside them. "I would say that we see even roles here exploring the market like data engineers, simulation engineers, digital twin analysts, and then also I think the skills to go towards scenario thinking and automation management is something that I do see as the next level," he said. Priorities for 2026 Looking ahead to 2026, Zaja shared that further scaling its digital twin model across more plants will be a top priority for FORVIA. "Second, I would say that we have to also follow [by] increasing the visibility of risk detection in the supply chain," he added. "Also, we have to further bring AI capabilities into "Second, I would say that we have to also follow [by] increasing the visibility of risk detection in the supply chain," he added. "Also, we have to further bring AI capabilities into our SNP process or even into the whole logistics system environment." He concluded: "The last and most important part is to develop our people in the skills of the supply chain and digital to really focus on that."