DRIVE PR – Data driven product realisation

Research · In the design office

Data-Driven Engineering Design

Getting the data from the rest of the life cycle into the methods and tools engineers use to design, analyse and verify products.

The design office is where most of a product's cost and behaviour is locked in, and where data from the rest of the life cycle arrives last, if at all. Engineers work with large, heterogeneous datasets from modelling, simulation, testing, production and service, but in most companies the data definitions and information models are embedded in legacy systems that were never meant to talk to each other.

DRIVE PR works on getting that data into everyday engineering: interoperable data definitions, predictive models trained on real production and operational data, and hybrid physics-based and AI models that give near real-time feedback while requirements change and complexity grows.

How do we define modular product platforms across integrators and suppliers without drowning in complexity?

Track: Data-driven modular and platform design

Next-generation modular and platform-based architectures that integrate several technologies and life-cycle aspects such as services, repair and circularity. Data-driven and AI-based methods reduce complexity, connect the modular structure to test and experimental data, and let engineers explore design alternatives systematically. Volvo Trucks and Autoliv share the question from its two sides: the integrator's and the subsystem supplier's.

Who: Autoliv and Volvo Group, with Chalmers researchers.

How do we design for situations we cannot test enough in reality?

Track: Design using synthetic datasets

Synthetic data, for example derived from video analysis, that represents complex real-world situations, so that new technologies can be designed, evaluated and optimised interactively in realistic use scenarios. The first example is occupant safety systems that adapt dynamically to where the occupant is and how the vehicle behaves.

How do we make designs robust to the variation the factory and the field will actually throw at them?

Track: Data-driven robustness

Feeding knowledge from manufacturing, inspection, repair and use back into early design decisions. By analysing the sources of variation across the life cycle, uncertainty and risk are managed before they become cost. Hybrid digital and experimental models, supported by digital twins and AI-based prediction, make design iterations fast and reduce technology risk.

Who: Saab and Volvo validate the results industrially. Chalmers geometry assurance researchers run the studies. RD&T Technology and PE Geometry carry the results into demonstrators and Learning Lab activities.

How do we design and validate products that only work together with other products?

Track: Systems-of-systems design of connected solutions

Design and validation of interconnected products that operate as systems of systems. Building on model-based systems engineering, realistic test and operational data are integrated for digital experimentation with coordinated, distributed systems. Ethical, legal and regulatory questions are part of the research. Volvo Construction Equipment plans to use field data to model and configure alternative product configurations, tested in the field at AstaZero with the Hugo test vessel from Engage Studios.

Who: Volvo, Engage Studios, AstaZero and the Chalmers Systems Engineering Design group.

The other impact areas