Design tasks need to be rescheduled and reprioritised frequently during product development. Inappropriate priority decisions generate rework; thus, the policy used to guide such decisions may have a significant effect on design cost and lead time. Generic priority rules provide easily implementable guidelines for task prioritisation and are theoretically effective for many planning problems. But can they be used in design processes, which include iteration, rework and changes? In this article, a discrete-event simulation model is developed to investigate priority policies in design. The model explores the combined effects of progressive iteration, rework and change propagation during design of interconnected parts in a product architecture. Design progression is modelled as an increase in the maturity of parts; rework and change propagation cause maturity levels in certain parts to reduce. Twelve product architecture models ranging in size from 7 to 32 elements are simulated to draw qualitative and general insights. Sensitivity of the findings to assumptions and model inputs is tested. Generally effective priority policies are identified, and their impact is shown to depend on the interconnectedness and organisation of product architecture, as well as the degree of concurrency in the design process.

In many engineering design contexts models are indispensable. They offer decision support and help tackle complex and interconnected design projects, capturing the underlying structure of development processes or resulting products. Because managers and engineers base many decisions on models, it is crucial to understand their properties and how these might influence their behaviour. The level of detail, or granularity, of a model is a key attribute that results from how reality is abstracted in the modelling process. Despite the direct impact granularity has on the use of a model, the general topic has so far only received limited attention and is therefore not well understood or documented. This article provides background on model theory, explores relevant terminology from a range of fields and discusses the implications for engineering design. Based on this, a classification framework is synthesised, which outlines the main manifestations of model granularity. This research contributes to theory by scrutinising the nature of model granularity. It also illustrates how this may manifest in engineering design models, using Design Structure Matrices as an example, and discusses associated challenges to provide a resource for modellers navigating decisions regarding granularity.

Models of products and design processes are key to interacting with engineering designs and managing the processes by which they are developed. In practice, companies maintain networks of many interrelated models which need to be synthesised in the minds of their users when considering issues that cut across them. This article considers how information from product and design process models can be integrated with a view to help manage these complex interrelationships. A framework highlighting key issues surrounding model integration is introduced and terminology for describing these issues is developed. To illustrate the framework and terminology, selected modelling approaches that integrate product and process information are discussed and organised according to their levels and forms of integration. Opportunities for further work to advance integrated modelling in engineering design research and practice are discussed.

Determining a suitable level of description, or granularity, for a product or process model is not straightforward, especially since granularity can manifest in multiple ways, but it is important to capture important elements in the model without building models that are too large to understand. This article investigates the implications of model granularity choices by simulating the design process of a diesel engine on different levels of detail, comparing the results and exploring ways to account for the differences. It uses two Design Structure Matrix (DSM) models for change prediction in a diesel engine at different levels of granularity to run simulations of the design process. Changes are a major source of rework and lead to frequent rescheduling of design tasks. The incremental nature of product development as well as design changes and their propagation complicate design process planning further. Process simulation may provide support in such contexts when it is based on an appropriate description of the product. The article shows that while coarse models can give an indication of likely process behavior, they miss potentially significant iteration loops.

This work studies operators mapping vector and scalar fields defined over a manifold M, and which commute with its group of diffeomorphisms Diff(M). We prove that in the case of scalar fields L p ω (M, R), those operators correspond to point-wise non-linearities, recovering and extending known results on R d . In the context of Neural Networks defined over M, it indicates that point-wise nonlinear operators are the only universal family that commutes with any group of symmetries, and justifies their systematic use in combination with dedicated linear operators commuting with specific symmetries. In the case of vector fields L p ω (M, T M), we show that those operators are solely the scalar multiplication. It indicates that Diff(M) is too rich and that there is no universal class of nonlinear operators to motivate the design of Neural Networks over the symmetries of M.Preprint. Under review.