AI-Based Demand Forecasting Brings New Planning Reliability to the Textile Industry

Artificial intelligence is becoming an important driver of transformation in the textile industry. As manufacturers face changing customer demand, seasonal fluctuations, and increasing pressure to improve efficiency, digital planning tools are creating new opportunities to make operations more accurate and reliable.

Reliability to the Textile Industry

To address these challenges, Fraunhofer IWU has developed an AI-powered demand forecasting solution for frottana Textil GmbH & Co. KG, the company behind the well-known MÖVE brand. The new system uses historical sales information and advanced analytics to support better sales and order planning while creating a foundation for future production optimization.

Traditional Planning Methods Create Challenges

Many textile and home textile manufacturers still rely heavily on conventional planning approaches. Excel spreadsheets, manual calculations, and employee experience remain common tools for forecasting demand and scheduling production.

This method can become difficult to manage, especially in medium-sized textile businesses where product ranges are large and demand patterns frequently change.

Seasonal peaks during spring collections, holiday periods, and festive shopping seasons create additional complexity. In many cases, teams recreate planning sheets every month for thousands of products, while some companies continue using handwritten records to support scheduling decisions.

Although businesses often possess years of valuable sales and production data, much of this information remains underutilized. Combined with labor shortages and the retirement of experienced employees, companies face the risk of losing valuable operational knowledge.

These limitations can result in planning uncertainty, higher manual workloads, production inefficiencies, and increased operational costs.

AI and Data Intelligence Transform Forecasting

Recognizing these challenges, frottana Textil GmbH & Co. KG partnered with Fraunhofer IWU and Logsol GmbH to introduce a digital forecasting solution powered by artificial intelligence.

The developed tool analyzes historical sales data and predicts future monthly demand while automatically identifying seasonal behavior and market trends.

Neural network technology plays a central role in the system by detecting complex relationships within sales data that traditional methods often overlook.

This creates a more reliable and transparent planning process that supports faster and more informed decision-making.

Strong Forecast Accuracy with Limited Historical Data

One of the most notable outcomes of the project was the forecasting quality achieved despite a relatively small dataset.

According to project findings:

• The AI model explained approximately 82.7 percent of sales fluctuations.
• The system successfully represented stronger month-to-month variations.
• Average forecast deviation remained around 38 units against average monthly sales of 340 units.

This represents an error rate of approximately 9 percent.

The performance was achieved using only four years of historical sales data and without including variables such as sales channels, regional performance, promotional campaigns, or special market disruptions including pandemic-related effects.

These results demonstrate that AI forecasting can deliver meaningful business value even when available data is limited.

Digital Planning with Human Expertise

A major advantage of the new solution is the automation of planning activities that previously required extensive manual effort.

The forecasting platform digitizes the planning process and provides employees with a transparent basis for decisions.

At the same time, the approach does not eliminate human involvement.

Employees continue to review, adjust, and enhance AI-generated forecasts using their own practical knowledge and business understanding.

This combination of machine intelligence and human expertise improves trust in the technology while supporting faster onboarding for new employees and reducing dependency on individual specialists.

Future Focus: Connecting Forecasting with Production

The next development phase aims to integrate demand forecasting directly into production planning.

By connecting forecast insights with manufacturing operations, textile companies may improve production sequencing, optimize batch sizes, and distribute manufacturing capacity more effectively throughout the year.

Fraunhofer IWU is also advancing broader digital initiatives through projects such as SmarMoTEX, which focuses on virtual material flow simulation to test production strategies and identify bottlenecks without interrupting active operations.

Additional innovation areas include image-based defect detection for weaving quality control and retrofitting older textile machines with modern sensor technologies to extend equipment life and improve operational performance.

As AI adoption accelerates across manufacturing, intelligent forecasting tools are expected to become an increasingly important foundation for building a more agile and competitive textile industry.