Modeling in the Design Multiverse

Sylvain Guérin, Salvador Martínez, Ciprian Teodorov · 2025

Read Modeling in the Design Multiverse by Sylvain Guérin, Salvador Martínez, Ciprian Teodorov (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-06530.txt.

Introduction

Design work that branches, merges, and forks across teams gets a new organizing idea from Sylvain Guérin and two co-authors in a 2025 software engineering paper. Real projects spawn many revisions and variants, yet modeling tools cannot manage that spread directly. Teams fall back on outside tools and habits.

The proposed Design Multiverse brings selected revisions and variants inside the modeling space itself. It stores snapshots of a design state, each made of several artifacts, so stakeholders can trace decisions, compare variants, and see how pieces depend on each other. Example uses include model product lines, where one design spawns many versions, and models that must evolve together with their own definitions. An implementation route based on model federation, linking separate model stores, is sketched.

Catalog record

CREATORSylvain Guérin, Salvador Martínez, Ciprian Teodorov
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE30 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-06530

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Modeling in the Design Multiverse Sylvain Guérin, Salvador Martínez, Ciprian Teodorov Abstract Real-world design processes often involve the evolution and divergence of design paths (by branching, revising, merging, etc.), especially when multiple stakeholders or teams operate concurrently and/or explore different alternatives for complex and heterogeneous systems. Unfortunately, this variability in time and space can not be directly managed in current modeling spaces but requires resorting to external tools and methodologies. In order to tackle this problem, we introduce the Design Multiverse . The Design Multiverse aims to integrate in the modeling space a selection of revisions and variants, representing snapshots of a design state composed of multiple artifacts. This enables stakeholders to seamlessly trace, analyze, and manage design decisions, system variants, and their interdependencies. Concretely, in this paper we present a conceptual definition of the Design Multiverse , discuss usage scenarios such as model product lines and model/metamodel co-evolution, and propose an implementation leveraging the model federation paradigm. I Introduction The INCOSE Systems Engineering Vision 2035 [ 1 ] outlines a future of systems that are resilient, sustainable, and adaptable throughout their lifecycle. Yet, current engineering practices fall short of this ambition. Modern cyber-physical systems are increasingly heterogeneous, spanning diverse disciplines, tools, and models, which makes it difficult to ensure coherence and consistency across the design process. Additionally, the lack of formal mechanisms to track the evolution of design decisions results in disconnected variant histories, unclear rationales, and difficulty adapting to changing requirements. Traditional approaches often treat design decisions and variants as static and isolated entities, limiting responsiveness to change. Critically, the absence of a temporal perspective on the design process impedes understanding of how past choices shape present configurations and future behaviors, which is an essential capability for building robust and adaptable systems. As a solution to the aforementioned issues, this paper introduces the Design Multiverse . The design multiverse is a conceptual, problem-framing framework, which permits understanding and addressing current problems in system design approaches and methodologies which are related to the lack of time and space awareness (e.g., the co-evolution of related artifacts, the co-existence of revisions and variants, etc.) of current modeling spaces. Concretely, the Design Multiverse integrates (e.g., makes co-exist) in the modeling space (e.g., model editors) a selection of revisions and variants, representing snapshots of a design state composed of multiple artifacts. In this setting, modelers are enabled to trace and analyze design decisions, system variants, and their inter-dependencies in a way that aligns with the future system engineering vision. We propose model federation [ 2 ] as an infrastructure for the realization of the Design Multiverse . Model federation is a multi-model management approach based on the use of virtual models and loosely coupled links. The models in a federation remain autonomous and are represented in their original technological spaces (and accessed through adapters), whereas virtual models (also called conceptual models) and links serve as control components used to present different views to the users and maintain synchronization through the use of automated behaviors. In the Design Multiverse scenario, links are used to maintain references between artifacts, revisions, and variants, adapters are used to present to the user coherent and usable multi-version multivariant models (e.g., by the creation of variant types) and behaviors to support tasks such as migration and co-evolution. The remainder of the paper is structured as follows. Section II introduces the Design Multiverse , followed by application scenarios in Section III . Section IV presents a prototypical realization based on model federation. Section V discusses related work. Section VI concludes providing future directions. II Design Multiverse We introduce the Design Multiverse as a problem-framing framework to understand and manage the complexities inherent in the evolution of design artifacts. Rather than assuming a linear trajectory of design activities, we acknowledge that real-world design processes often involve branching, merging, and divergence of design paths, especially when multiple stakeholders or teams operate concurrently, or when different alternatives are explored. The design multiverse captures these phenomena by modeling the design process as a partially ordered set of slices , each representing a snapshot of a design state, connected through design transitions . Fig. 1: Key notions in the Design Multiverse conceptual framework. The key notions are illustrated in figure 1 , which presents two interlinked design multiverses \mathcal{M}_{1} and \mathcal{M}_{2} . A slice is part of a design multiverse and is defined as a structured set of design artifacts (models, requirements, code elements, etc.), which are syntactic in nature. Tools and human interpreters operate over a slice to extract semantics, enabling reasoning, validation, and the formulation of design decisions that lead to the generation of new slices. The transitions between slices, capturing the local inter-slice design process, are deliberate evolutions that modify the design state based on rationale, analysis results, stakeholder input, or exploratory objectives. Formally, a design multiverse is a directed acyclic graph, where nodes are slices and edges are design transitions . A slice is generally labeled with a version. From a given slice, multiple new slices may be created, forming branches that represent: 1. Concurrent evolution by different teams; 2. Alternative responses to the same analysis results; or 3. Exploratory forks, where different hypotheses are evaluated in parallel. A slice is internally structured as rooted graph (with multiple roots potentially) where nodes are artifacts and edges internal references between artifacts (eg. conformance, usage, refinement, etc.). Design transition between slices can be accompanied by evolution links , which bind two successive versions of an artifact or an internal reference. Evolution links may carry semantic delta computation between the two versions (or a simple correspondence if artifact is unchanged), which eases coexistence and reasoning about multiple versions of a given artifact. Working with multiple design multiverses implies external dependencies . External dependencies are defined between artifacts from different slices and design multiverse. These fine-grained dependencies may be aggregated and expressed as slice external dependencies . The usual workspace of a design process is a combination slices from the required design multiverses. We call this workspace a composite slice , illustrated in figure 1 with version 2.0 of M1 and version 0.2 of M2 . A composite slice is syntactically closed if all external dependencies are resolved inside the composite slice itself. Together, these notions provide a structured foundation for modeling, analyzing, and reasoning about the inherently branching and evolving nature of real-world design processes. III Usage scenarios In the following, we describe two categories of scenarios that benefit from the Design Multiverse framing. Firstly, navigation scenarios , which involve slices in the same multiverse, and secondly co-evolution scenarios , which involve slices in at least two different multiverses. III-A Navigating through revisions and variants The multiverse enables modeling experts to seamlessly navigate between revision slices and its contained elements through evolution links. This navigation occurs in the modeling space so that parts of the model would appertain to a certain revision while other parts may have been navigated forward or backward in time. This way of navigating revisions in the modeling space permits the evaluation of hypothesis. These hypotheses may come in the form of formal constraints (or properties) which are (re)-evaluated upon navigation and allow modelers to answer questions such as: Does

Read the complete text