Intelligent scheduling with algorithms from MCP
Just imagine…
…you plan ten operations on one resource, which corresponds to 3,628,800 possible combinations!
We will be happy to do this for you!
Our aim is to provide mathematical optimisation in a practical form in order to solve complex industrial problems in an optimised way.
We develop algorithms to solve common industrial problems and make them available in an algorithm library: set-up time minimisation, batch optimisation, resource allocation and many other tasks are our playground. And with the APS AI (Algorithm-as-a-Service) concept, you can easily access our solutions via a REST interface – modern and uncomplicated.
We rely on close cooperation with research to close the gap between academic expertise and actual, practical application in industry. As the main investor, we have successfully operated a laboratory of the Christian Doppler Research Association together with Bosch and Ximes. The result is not only a large number of scientific publications, but also practically applicable methods from the field of artificial intelligence.
We are convinced that the next step for ERP, MES and PPS software is AI-supported automatic plan optimisation. With APS AI, we are now making this possible – for modern and efficient production without waste.
Let’s talk about your planning challenges:
Research network
We are funding a doctoral position at the Doctoral College “Innovative Combinations and Applications of Artificial Intelligence and Machine Learning” (iCAIML). This strengthens our connection to the Institute of Logic and Computation at TU Vienna and enables us to keep our finger on the pulse of the exciting topics of hyperheuristics and automated algorithm selection.
Our Christian Doppler Laboratory was set up at the Institute of Logic and Computation at TU Vienna from 2017 to early 2025. There, together with scientists, we conducted targeted basic research into the development of new algorithms and the use of artificial intelligence in production planning.
Together with our partners Bosch and Ximes, we pioneered research into the practical application of artificial intelligence methods for complex problems within production planning.
Together with the Karlsruhe Institute of Technology (KIT), we are researching methods for making industrial processes more flexible in terms of energy. The aim is to develop AI-supported methods that make energy-intensive production more grid-friendly and sustainable. The collaboration combines our many years of expertise in APS with KIT’s cutting-edge research in the field of energy system design. As an implementation partner, we bring research results into industrial practice – for sustainable production.
Extract from research areas
- Nonlinear dynamic systems
- Operations Research
- Multicriteria optimisation
- Metaheuristic optimisation methods
- Business Process Management
- Artificial Intelligence
Extract from research areas
- Stochastic modelling of production and logistics systems
- Configuration of flexible flow production systems, capacity planning
Extract from research areas
- Human factors: tactical and operational planning in production
- Industry 4.0: planning of networked systems and optimisation of data collection and data maintenance
- Decision analysis
- Project management
Our algorithm solutions
Setup Time Optimization
The problem is to create a production plan for several lines running in parallel (parallel machine scheduling), with set-up times (changeover times) occurring between successive orders. Differences between machines in terms of set-up and processing times must be taken into account. Shift calendars, secondary resources and dependencies between orders can also play a role.
The aim is to minimise the total amount of set-up work while taking due dates into account.
Production systems with sequence-dependent and (strongly) fluctuating set-up times on one or more lines.
Food Production Scheduling
This optimisation algorithm was developed to solve the planning problem of a typical food and beverage production scenario. The scenario includes orders for the production of intermediate products and packaging. Between these production phases there are intermediate warehouses, each of which is subject to a maximum capacity limit. Restrictions or priorities can be taken into account to determine which intermediate product may or should be stored in which intermediate storage facility.
The optimum planning sequence for the production and/or packaging area is calculated taking set-up matrices and requirement dates into account.
Manufacturing with multi-stage production and intermediate storage as an intermediate step. Frequent areas of application are production in the food & beverage industry and the pet food industry, as well as the process industry (chemicals, building materials, etc.).
Batch Optimization
The Oven Scheduling Problem (OSP), is a parallel batch scheduling problem. Orders must be scheduled for one of several ovens and can be processed simultaneously with other orders in a batch if they have compatible requirements. When scheduling the orders, various constraints regarding the suitability and availability of the ovens, the release dates of the orders, the set-up times between batches and the oven capacities must be taken into account.
Reduction of the oven running time by forming suitable “batches”, while at the same time taking into account demand deadlines and minimising the throughput time.
For example, during heat treatment in the manufacture of electronic components.
Production Leveling
Production levelling involves an even distribution of the production volume (total and per product) over individual periods. A period can be, for example, a shift, a day, a week or a month. Production levelling is an important part of implementing the Heijunka principle.
Levelling of production quantities across periods and product types, while at the same time taking into account priorities/demand dates and secondary restrictions.
Production levelling is often necessary in electronics manufacturing, for example, in order to achieve a balanced product mix on a daily basis and to enable optimum utilisation of production capacities in scheduling. In addition, levelling or smoothing often plays a decisive role in long-term capacity/production planning.
Resource Assignment
The feasibility of a production plan usually depends on other resources, such as production aids or interim storage capacities. When it comes to the actual allocation of resources, the number of possibilities very quickly exceeds what a human can manage. Mathematical optimisation can achieve much better results in seconds than a person can achieve through lengthy deliberation. Solution using simulated annealing (metaheuristic framework).
Secondary resources (e.g. boilers, tanks, tools, etc.) required as part of the production process are allocated to the orders on the basis of an existing production plan.
Production with production aids or intermediate storage in containers.
Employee Scheduling
This algorithm is used to allocate the existing available capacity (employee availability) to the capacity requirements, taking into account various restrictions and parameterisable optimisation targets. The algorithm is called up from MCP Workforce Management, for example. The employees are allocated to work centres.
Distribute employees to workstations within a time period so that the planned orders can be processed in the best possible way.
There are various optimisation targets that can be combined with each other through prioritisation and weighting (priority of employees or workstations, deployment of best-qualified employees, etc.).
Employee scheduling at shift level.
Automatic Parameterization of the MCP Weighting Rule (Machine Learning)
Using a trained model, an artificial intelligence suggests a weighting of the individual optimisation parameters based on the current planning situation.
Improved planning results in a dynamic environment through AI-supported weighting of target factors.
Industrial production whose planning depends on several optimisation factors, whereby the importance of the factors in relation to each other changes repeatedly over time (e.g. fluctuating capacity utilisation, seasonal effects).
Paint Shop Scheduling
A large number of parts have to be painted every day in paint shops in the automotive supply industry or in machine and equipment construction. The product carriers are transported through the plant via conveyor systems. Planning takes place in rounds. Many factors can increase the complexity of the planning task: Colour sequences; minimum and maximum block sizes; technical properties of the product carriers; cleaning processes; restrictions on hanging and unhanging; parallel painting booths; and much more.
Minimisation of colour changes and costs, taking into account due dates and technical and procedural restrictions.
Painting systems in the automotive industry, electronics industry and metal processing: booth painting systems, automated painting systems, powder coating systems.
Artificial Teeth Scheduling
Tooth production takes place in round systems. The base material is injected into metal moulds and processed into a raw tooth in a multi-stage process within a single pass in the system.
Efficient loading of the system, taking into account the delivery deadlines and the moulds as a limited production resource.
In addition to tooth production, very similar planning challenges can be found in rotary systems in the food, cosmetics, pharmaceutical, chemical and packaging industries.