Imagine you're a film or game studio and every time you start a new project, your team rebuilds 3D characters, animations, and visual effects from zero — even though you made similar ones last year. SAUCE built smart tools that automatically tag, search, transform, and adapt existing digital content so it fits new productions without starting over. Think of it like a smart library that not only finds what you need but reshapes it to fit your current project. The tools cover everything from motion-capture editing to crowd scene generation to virtual production environments.
By the numbers
EUR 3,999,375
EU funding for smart asset reuse R&D
9
consortium partners across the project
6
countries represented in the consortium
10
working prototype demonstrations delivered
30
total project deliverables completed
4
industry partners in the consortium
3
search modes in the Smart Search Prototype
The business problem
What needed solving
Creative content studios — in film, games, broadcasting, and advertising — waste enormous budgets recreating digital assets that already exist somewhere in their archives or in previous productions. Finding, adapting, and reusing 3D models, animations, and visual effects across different projects and production pipelines is manual, slow, and technically painful. This means higher costs, longer production timelines, and creative teams stuck doing repetitive work instead of innovating.
The solution
What was built
The project built 10 working prototypes: a smart asset storage system (cloud + local), a search engine with three search modes for digital assets, a semantic labelling toolbox using deep learning, motion-capture editing tools (path alteration, style transfer, collision solving), animation splicing tools, a virtual production toolkit, crowd scene synthesis with quality metrics, transitional animation generation, and a transcoding mechanism for cross-pipeline compatibility. All 30 deliverables were completed by project end in 2020.
Audience
Who needs this
VFX and post-production studios managing large asset libraries across multiple productionsGame development studios building character-heavy or open-world titlesVirtual production companies and broadcasters needing real-time content adaptationAnimation studios looking to automate crowd scenes and motion-capture editingMedia technology companies building content management and asset pipeline tools
Business applications
Who can put this to work
Film & Visual Effects Production
any
Target: VFX studios and post-production houses
If you are a VFX studio dealing with ballooning production costs because every project demands new 3D assets built from scratch — this project developed a smart asset store with automated search and semantic labelling that lets you find, adapt, and reuse existing assets across productions. The consortium of 9 partners across 6 countries built 10 working prototypes including smart search and asset transformation tools.
Video Game Development
mid-size
Target: Game studios producing open-world or character-heavy games
If you are a game studio spending months hand-animating characters and crowd scenes for every new title — this project developed tools for splicing animation clips, editing motion-capture data, and automatically synthesizing crowd scenes with quality metrics. These tools let you generate smooth transitional animations from existing assets rather than creating each one manually.
Virtual Production & Broadcasting
enterprise
Target: Broadcasters and virtual production studios
If you are a broadcast or virtual production company struggling to keep up with demand for real-time content — this project developed a virtual production prototype toolkit and light-field capture technology that makes digital assets adaptive to different production environments. The transcoding mechanism ensures assets work across different pipelines and remain usable as technology evolves.
Frequently asked
Quick answers
How much could this actually save on content production costs?+
The project objective states these tools enable 'a vast reduction of costs and increases in efficiency' in content production by enabling reuse instead of recreation. Specific cost savings percentages were not published in the available data. The real savings come from not rebuilding assets for each new project — a studio producing multiple titles per year would see compounding returns.
Can these tools work at industrial production scale?+
The project delivered 10 demonstrated prototypes across 30 total deliverables, including a cloud-and-local asset storage system designed for delivery regardless of user or asset location. The smart search prototype supports three search modes (property, tags, and descriptors). These were demonstrated at industry events via the Final Showcase Demonstration, but full production-scale deployment would require further integration work.
What about intellectual property and licensing?+
SAUCE was funded as a Research and Innovation Action (RIA) under Horizon 2020 with EUR 3,999,375 in EU contribution. RIA projects typically allow partners to retain IP on their contributions. With 4 industry partners and 5 universities in the consortium, licensing arrangements would need to be negotiated with the specific partner holding the relevant technology.
How does the smart search actually work?+
The Smart Search Prototype allows users to find assets using three modes: searching by property, by tags, and by descriptors. It also includes a 'related assets' feature that surfaces indirectly matching content. The underlying technology uses deep learning and semantic labelling on both 2D and 3D data for automated classification.
What animation tools were specifically built?+
The project delivered tools for editing motion-capture data (path alteration, style transfer, collision solving), tools for splicing animation clips together using motion graph methods, crowd scene synthesis with quality evaluation metrics, and transitional animation generation based on semantic constraints. All were demonstrated as working prototypes.
How long would integration into existing pipelines take?+
The project ran from 2018 to 2020 and produced prototype-level tools. The Virtual Production prototype toolkit and Transcode Mechanism were specifically designed to work with existing media production pipelines. However, as these are research prototypes with 10 demos delivered, integration into commercial production environments would require engineering effort beyond what was demonstrated.
Is there ongoing support or development?+
The project ended in December 2020 and is now closed. The coordinator is Universidad Pompeu Fabra in Spain. Post-project development would depend on whether individual partners continued commercializing their specific tools. The project website at sauceproject.eu may contain information about follow-up activities.
Consortium
Who built it
The SAUCE consortium brings together 9 partners from 6 countries (Switzerland, Czech Republic, Germany, Spain, Ireland, UK), with a nearly balanced split of 4 industry and 5 university partners — giving it a 44% industry ratio. This mix is strong for a creative technology project: the universities (led by coordinator Universidad Pompeu Fabra in Spain) provide deep learning and computer vision research, while the 4 industry partners including 2 SMEs bring real production-world requirements. The EUR 3,999,375 EU investment across this team produced 30 deliverables including 10 working demos, suggesting productive use of resources. For a business considering these tools, the presence of real creative industry companies in the consortium means the prototypes were built with actual production needs in mind, not just academic curiosity.
Universidad Pompeu Fabra (Spain) coordinated the project — contact their research technology transfer office for licensing and collaboration inquiries.
Want to explore how SAUCE's smart asset reuse tools could cut your content production costs? SciTransfer can connect you directly with the right research team. Contact us for a tailored brief.
