Hettich

Hettich, one of the world's largest furniture fittings manufacturers, operates complex warehouse logistics that must balance predictability against extreme variability. The rise of e-commerce introduced a steady stream of short-notice orders requiring next-day delivery, pushing daily order volumes into unpredictable territory. Across six distinct warehouse areas — each with its own functional requirements — this volatility translated directly into staffing uncertainty. A structural shortage of skilled logistics workers made flexible short-term hiring impractical, compounding the planning gap. Without reliable demand visibility beyond a few days, planners had no basis for proactive resource allocation, leaving operations chronically reactive and exposed to service failures under peak load.

In collaboration with pacemaker.ai — a thyssenkrupp subsidiary specializing in AI-driven supply chain optimization — Hettich deployed a cloud-based SaaS forecasting solution powered by machine learning algorithms. The platform was configured to cover all six of Hettich's distinct warehouse areas simultaneously, delivering demand forecasts at daily, weekly, and monthly time horizons within an 84-day forward-looking window. Rather than requiring a disruptive system migration, the solution integrated through an automated daily data exchange that refreshes forecasts each morning. Forecasts are accessible via pacemaker.ai's purpose-built frontend and can be exported to Excel, enabling adoption within Hettich's existing IT landscape. The system is self-learning, continuously improving predictions as additional operational data accumulates.

The deployment achieved forecast accuracy exceeding 90% across an 84-day planning horizon — a significant advance for a logistics operation where reliable visibility beyond a few days had previously been out of reach. This performance spans all six warehouse areas, giving planners consistent, comparable signals across the entire operation. With precise work-requirement forecasts available per area and per time horizon, Hettich can now make proactive staffing decisions despite tight labor market conditions.

• >90% forecast accuracy at an 84-day horizon
• 6 warehouse areas covered under a single unified model
• Daily, weekly, and monthly forecast outputs address distinct planning cycles
• Reduced dependence on reactive short-notice hiring in a constrained labor market

• Warehouse demand forecasting must produce multi-horizon outputs — daily, weekly, and monthly signals serve different planning cycles and cannot be substituted for one another.
• Unifying all functional warehouse areas under a single forecasting model prevents siloed planning and ensures comparable resource signals across the operation.
• Offering forecast exports in familiar formats like Excel alongside a dedicated UI drives adoption without forcing changes to established IT workflows.
• A self-learning model that ingests daily operational data maintains and improves accuracy over time without requiring manual retraining cycles.
• In labor-constrained logistics environments, forecasting converts reactive hiring into proactive scheduling — shifting from variable recruitment cost to predictable headcount planning.