Research
AIDIP project: AI and Humans together to improve manufacturing overall equipment efficiency
News
July 9, 2025
Every industrialised country generates a significant portion of its wealth from the manufacturing sector, particularly in complex manufacturing, which is typically called the high-mix-low-volume (HLMV) sub-segment, which is typically referred to as the High-Mix Low-Volume (HMLV) sub-segment, constituting 20 to 70% of it, depending on the country or industry. The AIDIP project, funded by EIT Manufacturing in 2024 and involving a consortium of companies – LexaTexer, Solenis, Beko Europe and Cefriel – aimed at developing an advanced production planning support system, based on the construction of a Data Space and powered by Artificial Intelligence. AIDIP was focused on complex manufacturing, involving numerous machines that are approximately 4 to 6 times the size of a desk.
LexaTexer deals with the creation of results that can be operationalised. To achieve this goal, it has built software platforms for onboarding, integrating, managing, securing, analysing, and deploying data-driven solutions. These solutions solve real-world problems across industries, mostly focused on manufacturing and intelligence use cases. Their products automate the process of predictive analytics, thus drastically reducing the time to deliver results. LexaTexer played a significant role in the AIDIP project, contributing to the initiative’s both technical and commercial aspects. In particular, LexaTexer focused on the definition of advanced AI algorithms and the logic for production planning. LexaTexer’s sales team is responsible for reaching out to prospects and proposing the AIDIP solution to customers and third-party tech companies, in order to introduce the AIDIP system to the high-mix-low-volume (HMLV) manufacturing market through a strategic commercial plan.
LexaTexer deals with the creation of results that can be operationalised. To achieve this goal, it has built software platforms for onboarding, integrating, managing, securing, analysing, and deploying data-driven solutions. These solutions solve real-world problems across industries, mostly focused on manufacturing and intelligence use cases. Their products automate the process of predictive analytics, thus drastically reducing the time to deliver results. LexaTexer played a significant role in the AIDIP project, contributing to the initiative’s both technical and commercial aspects. In particular, LexaTexer focused on the definition of advanced AI algorithms and the logic for production planning. LexaTexer’s sales team is responsible for reaching out to prospects and proposing the AIDIP solution to customers and third-party tech companies, in order to introduce the AIDIP system to the high-mix-low-volume (HMLV) manufacturing market through a strategic commercial plan.
Overall equipment efficiency through automated scheduling
High-mix-low-volume manufacturing requires frequent reconfiguration of machines on the shop floor, often multiple times a day. These machines contain tools that need to be reconfigured several times daily, taking anywhere from 30 minutes to 2.5 hours. During this timeframe, production is halted, leading to idle time and potential financial loss.
Since there are no formulas to find an optimal schedule, optimising this process typically requires human teams to generate detailed shop floor schedules dictating specific items to be produced at particular times, on designated machines, and by assigned teams. This scheduling task is challenging, due to its NP-hard mathematical nature, requiring extensive time, often two to three days.
Since there are no formulas to find an optimal schedule, optimising this process typically requires human teams to generate detailed shop floor schedules dictating specific items to be produced at particular times, on designated machines, and by assigned teams. This scheduling task is challenging, due to its NP-hard mathematical nature, requiring extensive time, often two to three days.
“Generating these schedules – Dr. Günther Hoffmann explains – involves daily meetings and discussions, adding complexity and consuming considerable time. Additionally, unexpected issues such as machine failures, missing raw materials, absent team members, and urgent orders from a high-priority customer also necessitate frequent rescheduling. And after each of these events, people have to restart the scheduling operations to optimise the process.”
At the same time, there is an additional layer of complexity because people talk to different roles with different personas. Different roles within manufacturing activities imply varied KPIs, at least two to three different layers with different KPIs, thus complicating schedule optimisation. Also, this project aims at adding another layer of optimisation, because it focuses both on standard KPIs like utilisation and deadline compliance, and on optimisations for emissions, too.
“This project – Dr. Günther Hoffmann says – aims at enhancing schedule generation by incorporating emission reduction alongside with standard KPIs like utilisation and deadline compliance. The automated system developed by the AIDP project can quickly produce reproducible and optimised schedules, adapting to changing KPIs. For example, it can prioritise emission profiles one day and the deadline compliance the next.”
Implementing automated scheduling significantly improves various manufacturing KPIs, including improved utilisation, downtime reduction, output increase, and overall equipment efficiency (OEE). Although automated scheduling may initially face resistance from schedulers and operators, it is essential to note that human oversight is always involved, ultimately turning out to be crucial in the creation of better schedules efficiently. “This is not merely a scheduling tool; it is a simulation tool designed to assist schedulers”, Dr. Günther Hoffmann highlights. “It provides support, which is crucial for improving scheduling efficiency. Hence, schedulers will not face unemployment but will instead enhance their productivity through the utilisation of this tool.”
“This project – Dr. Günther Hoffmann says – aims at enhancing schedule generation by incorporating emission reduction alongside with standard KPIs like utilisation and deadline compliance. The automated system developed by the AIDP project can quickly produce reproducible and optimised schedules, adapting to changing KPIs. For example, it can prioritise emission profiles one day and the deadline compliance the next.”
Implementing automated scheduling significantly improves various manufacturing KPIs, including improved utilisation, downtime reduction, output increase, and overall equipment efficiency (OEE). Although automated scheduling may initially face resistance from schedulers and operators, it is essential to note that human oversight is always involved, ultimately turning out to be crucial in the creation of better schedules efficiently. “This is not merely a scheduling tool; it is a simulation tool designed to assist schedulers”, Dr. Günther Hoffmann highlights. “It provides support, which is crucial for improving scheduling efficiency. Hence, schedulers will not face unemployment but will instead enhance their productivity through the utilisation of this tool.”
Data Space for Sustainability
One of the main KPIs that are often overlooked in contemporary production planning is the sustainability impact, particularly regarding logistics. This impact can be significant and influence the final production emissions.
“To assess this aspect as well, Cefriel developed a Data Space that gathers information from potential suppliers, thus allowing the collection of information about the possible material alternatives that are available at any time. A dedicated platform calculates the sustainability impact of each material, considering not only their production emissions but also the logistics emissions to deliver the material to the production site“, Andrea Villa, Smart Industry Research Lead, Cefriel, explains.
In the optimisation phase, the scheduler algorithms consider the sustainability impact KPI and propose various plans based on the desired emissions for a specific product. Last but not least, in this project, technical skills for supply chain and plant systems development are as relevant as coordination efforts for requirement collection and integration. “Cefriel’s role in coordination and management, as well as its collaborative technical development, was crucial for AIDIP’s success, addressing uncertainties and ensuring smooth collaboration”, Günther Hoffmann states.