Häufig gestellte Fragen
What is an APS system and how does it differ from MRP?
APS (Advanced Planning and Scheduling) is a class of planning systems that optimise production, material and capacity planning simultaneously while taking real-world constraints into account. The key difference from classic MRP lies in how capacity is treated: MRP generally calculates against infinite capacity and delivers dates that often prove unrealistic in practice, whereas APS factors in finite capacities such as machine hours, staff qualifications, setup times and tool availability. As a result, an APS system generates not just demand proposals but feasible, concretely scheduled and sequence-optimised plans. Mathematically, heuristics, constraint solvers and optimisation algorithms are used for this.
Do I need APS in addition to the ERP?
That depends heavily on manufacturing complexity. For complex engineer-to-order or high-variant manufacturers with multiple bottlenecks and scarce resources, a dedicated APS delivers considerably more realistic dates than the ERP's own materials planning. For simple series production with low variant diversity, by contrast, the demand planning integrated into many ERP systems is often sufficient. In practice, the exact design varies by industry, company size and the customisation depth of the specific ERP setup, which is why an individual assessment of your own planning problems is recommended.
How do APS, ERP and MES differ from one another?
The three systems cover different levels and complement each other. An ERP system forms the transactional basis with master data, orders, inventories and postings; APS builds on this to handle forward-looking, optimising planning against finite capacities; and an MES controls execution and feedback on the shop floor in real time. Ideally a closed control loop emerges: the ERP supplies orders and material data, APS creates a feasible plan, the MES executes it and reports actual data back, whereupon APS replans in the event of deviations. APS thus replaces neither ERP nor MES but adds the capability for constraint-based detailed scheduling.
What does an APS system cost?
Costs depend on functional scope, modelling depth and integration effort and can only be roughly categorised. In the mid-market, classic licence costs frequently range from around 50,000 to 300,000 EUR depending on the solution and module scope, typically plus implementation costs of a similar or higher magnitude, while simple, lightweight cloud-based subscriptions are already available from a low two- to three-digit euro amount per user per month. Importantly, pure licence fees usually account for only the smaller part of total costs, with the majority going to implementation, customisation, training and data migration; for enterprise solutions, two to three times the software costs are often budgeted for implementation. The ranges quoted are reference values and should be replaced by concrete quotes for your own use case.
Which APS vendors and systems are on the market?
The market ranges from modules integrated into ERP suites to specialised standalone systems. In the enterprise segment, SAP (with PP/DS and Integrated Business Planning), Siemens Opcenter APS (formerly Preactor), DELMIA Quintiq/Ortems, the Japanese Asprova and platforms such as Kinaxis, Blue Yonder and o9 are widespread, among others. In the mid-market there are also solutions such as PlanetTogether and the ERP vendors' own detailed-scheduling and planning-board modules. The market information reflects the state of ongoing research; a selection should always be made along your own requirements for modelling depth, ERP integration and performance, not solely by vendor size.
What prerequisites are needed for a successful APS implementation?
Data quality is decisive: unmaintained routings, incorrect setup times, outdated capacity data or incomplete shift and calendar data lead to unrealistic plans, regardless of how powerful the algorithm is. A step-by-step approach is recommended — first clean master data, then a realistic planning model and only afterwards optimisation towards a few clearly defined targets such as on-time delivery, throughput time or setup effort. A common mistake is to start with too many target variables at once, as overloaded models produce plans that are hard to comprehend and that planners do not trust. Industry sources often cite a duration in the order of about six to twelve months for APS projects, which can be shorter for pilot approaches and deviate significantly where complexity is high.