KIAS

AI-supported quotation calculation in special machine construction

problem

Mechanical engineering at the Saar has been one of the structurally shaping industries for many years. Woll Maschinenbau produces customer-specific special machines and production systems for customers in the automotive, food and medical industries at its Saarbrücken site. Two success-critical processes are of enormous importance for the acquisition and implementation of projects in special machine construction. These are the processes of quotation preparation and recalculation of projects. The preparation of the offer is carried out by project managers with the help of employees from individual departments. It should be noted here that the number of requests for tenders is increasing and, due to competitive pressure, many times as many tenders have to be drafted as are ultimately commissioned. This leads to increased ‘price and volume pressure’ on the bids to be issued. Particularly in special machine construction, there is the dilemma between an offer that is as precise as possible and the submission of an offer that is only roughly calculated.

objective

The aim of the research project KIAS is the optimization of existing production and company processes using artificial intelligence at the company Woll Maschinenbau GmbH. The focus of the developments in the research project lies in the area of internal processes of quotation preparation (from quotation preparation to post-calculation) for the project planning of plants in special machine construction (construction, production, assembly and commissioning as well as service). Through innovations, the aim is to make the ongoing processes in quotation creation and project tracking more accurate, faster and more cost-efficient, thus ensuring a competitive advantage.

approach

Phase 1: In the first phase, data is collected using the app developed in the course of the research project NeWIP on the shop floor.

Phase 2: Taking into account the nature and ‘design’ of the data collected, the second phase serves to develop a data model that enables the structured storage and further processing of historical and current process data in the course of the quotation calculation.

Phase 3: This phase describes the procedure for taking into account uncertainty and other uncertainties related to the non-specific customer requirements in early project phases and the measurement uncertainties in data collection.

Phase 4: The subsequent fourth phase serves to develop an adapted method for quotation calculation with regard to the cost calculation scheme used at Woll Maschinenbau and the conventionally used approaches.

Phase 5: Based on the developed calculation scheme and the method for taking into account uncertainty, the calculation algorithm and a hybrid method for implementing the calculation are now being developed. This hybrid calculation method combines basic approaches to cost estimation and takes into account different cases depending on the degree of similarity of the historical data to the new product.

Phase 6: This phase deals with the detailed conception and development of the human-machine interaction including the actors, the system components involved and the exact sequence of the computer-aided cost calculation.

The research project is divided into five work packages:

AP1: In step 1, an analysis of the current offer process and the post-calculation is carried out.

AP2: In step 2, the standards and models for the description of data and future processes are developed.

AP3: The aim of step 3 is to collect, interpret and process data based on the standards, models and ontology defined in AS2.

AP4: In step 4, the analysis, evaluation and further development of deterministic, stochastic and parametric calculation models, AI algorithms (predictive costing) and the development of a human-machine interface takes place.

AP5: In AS 5, the first prototypes of the system to be developed for AI-supported quotation calculation in special machine construction are created.

Contact person: Martin Karkowski, M.Sc.

Duration: 01/12/2020 – 30/11/2022