Both CHiPS and NoMADS are rooted in computational mathematics — structure-preserving algorithms and spectral operator methods — reflecting MathWorks' core MATLAB platform expertise.
THE MATHWORKS LIMITED
Commercial software company (MATLAB/Simulink) contributing numerical computing and machine learning tool expertise to European mathematical research consortia.
Their core work
MathWorks is the company behind MATLAB and Simulink — the industry-standard platforms for numerical computing, algorithm development, and model-based design used by engineers and scientists worldwide. Their H2020 participation reflects their role as an industrial software partner in academic research consortia, contributing computational tools, algorithmic expertise, and staff exchange to European mathematical and data science research networks. In both projects, MathWorks operated as a practitioner bridge: bringing industrial software engineering discipline into research environments focused on advanced mathematical methods. Their engagement with nonlocal methods, spectral analysis, and machine learning aligns directly with MATLAB's core application domains.
What they specialise in
NoMADS (2018–2023) explicitly lists machine learning, data science, and efficient algorithms among its focus areas, matching MathWorks' commercial product lines.
NoMADS keywords include finite weighted graphs, spectral operator decomposition, and nonlocal methods — advanced mathematical techniques for processing irregular data structures.
NoMADS lists biomedical imaging and point cloud processing as application domains, indicating MathWorks contributed computational tools to applied medical and 3D data analysis tasks.
How they've shifted over time
Their first project, CHiPS (2016–2019), left no recorded keywords, suggesting MathWorks played a supporting or tooling role without driving the intellectual agenda. By their second project, NoMADS (2018–2023), a clear and rich keyword profile emerged — nonlocal methods, spectral decomposition, machine learning, biomedical imaging — signalling a more substantive research engagement aligned with MATLAB's evolving product directions. The shift tracks the broader industry move: from pure numerical computing toward data-driven and graph-based methods that are now central to the modern MATLAB ecosystem.
MathWorks appears to be deepening its engagement with academic research in graph signal processing and applied machine learning — areas where MATLAB competes with Python ecosystems — suggesting future collaborations will likely center on data-driven mathematical methods.
How they like to work
MathWorks participates exclusively as a consortium partner, never as coordinator — consistent with a large commercial company that joins academic-led MSCA-RISE networks to provide industrial relevance and staff exchange capacity rather than to lead research programmes. With 31 unique partners across 13 countries from just two projects, they operate in large, internationally distributed consortia. This profile suggests they are accessible as industrial partners but unlikely to anchor or coordinate a consortium.
MathWorks has built a surprisingly broad network for just two projects — 31 partners across 13 countries — reflecting the MSCA-RISE model of multi-institutional staff exchanges. Their network is European in composition but with likely connections to their global academic user base.
What sets them apart
MathWorks is not a research organisation — it is the commercial software vendor that most of the research community already uses daily. That is precisely what makes them an unusual consortium asset: they bring tool adoption pipelines, direct product feedback loops, and industrial validation that purely academic partners cannot offer. For any consortium working on computational mathematics, signal processing, or machine learning that needs a pathway from algorithm to deployable software, MathWorks offers a connection that is hard to replicate with another partner.
Highlights from their portfolio
- NoMADSThe richest and most recent of their two projects, NoMADS ran five years (2018–2023) and generated a dense keyword set spanning nonlocal mathematics, machine learning, biomedical imaging, and point cloud processing — the clearest signal of where MathWorks' research engagement is heading.
- CHiPSMathWorks' entry into H2020 via a structure-preservation mathematics project, notable as an early signal of their strategy to embed their platform within European fundamental research consortia through MSCA-RISE staff exchanges.