Design agent transforming customer needs to optimal machine concepts

Challenge

Machine builders are constantly striving for better performance, lower costs and more energy-efficient machines. Yet new designs often build on existing machine concepts and familiar components. As a result, innovations tend to remain incremental, whilst the underlying machine concept is not necessarily optimal.

However, a large part of the final machine performance and costs is already determined during the conceptual design phase. Factors such as energy consumption, component cost, accuracy and total cost of ownership are strongly influenced by the chosen mechanical architecture and motion concepts. Despite this potential, companies often avoid fundamental redesigns because they involve high development costs, risks and longer time-to-market.

Today, there is a lack of efficient tools to automatically generate, evaluate and optimise alternative machine concepts. Machine designers therefore rely heavily on their own experience and familiar solutions, meaning that potentially higher-performing or more energy-efficient concepts are not sufficiently explored. Furthermore, existing optimisation processes still require many manual intermediate steps and support only a limited range of mechanical configurations.

Project goals

This project (ODA-CONCEPT_SBO) aims to develop an intelligent design agent that automatically translates customer requirements into optimised machine concepts for positioning and motion applications.

The solution will automatically generate machine concepts based on available components, analyse possible mechanical configurations and optimise the highest-performing variants in terms of energy consumption, drive forces, cost and performance. In doing so, the project aims to provide virtual insight into potential improvements compared to existing benchmark concepts.

The project builds on previous research findings regarding concept optimisation and focuses on automating the entire design workflow, from concept generation to modelling and optimisation. This involves integrating CAD models, geometric algebra and optimisation algorithms into a single, consistent design environment.

In the longer term, the project aims to support machine builders in exploring innovative machine concepts more quickly, reducing design complexity and realising higher-performance, more energy-efficient and cost-effective machines.