SW-MOTECH
Commissioned development · MVP
SW-MOTECH wanted to use future product demand in supply-chain planning. With a large product range, estimating demand manually for every item becomes costly. Short-notice delivery requirements also affect costs throughout the supply chain.
IAP developed a forecasting AI that analyses historical sales and combines its predictions with an understandable assessment of forecast errors. Planners can see which forecasts offer useful guidance and which products need closer attention.
Development took account of the customer’s product categories. IAP selected forecasting methods for the specific task and aligned evaluation with each category and the business objective. Historical sales provided the input for the forecasts.
Evaluation goes beyond a single average error figure. The solution examines error distributions and translates them into grades from 1 to 5. Mirrored-density plots make it possible to examine the distribution of forecast errors. This helps users see how a prediction compares with the errors observed in the evaluation.
The application reads a standardised data format defined for the customer. Staff can select an item and view its historical sales alongside the forecast, displayed uncertainty bounds and error grade. Forecasts, plots and processing logs can be saved.
For important products or poor error grades, the interactive view supports closer investigation. Knowledge of exceptional events, such as an unusually large individual sale, helps users interpret unusual patterns. This human-in-the-loop approach combines automated calculation with the experience of the people responsible for planning.
IAP delivered the solution as a Docker container with an interface accessible through a browser. Calculations run locally and use multiple processor cores. Once set up, the application can be used without programming knowledge.
The preceding Proof of Concept (PoC) showed a reduction in mean absolute percentage forecast error. It also established the basis for customer-specific method selection and assessment of expected errors. This work led to the decision support MVP for supply-chain planning. Jürgen Swora describes SW-MOTECH’s assessment of its value in his complete original statement below.
SW-MOTECH
The custom-made forecasting AI developed by IAP-GmbH provides important insights into future demands for our products that allows us to reduce overall supply chain costs. The acquisition of the competitor off-the-shelf software would have been ten-times more costly with no better forecasts.
General Manager
Original statement · English
In a free, non-binding 20-minute introductory discussion, we clarify your objective, starting point and a useful next step.