InFuse · Project

Open-Source Sensor Fusion Software for Space and Industrial Robotics

digitalTestedTRL 5

h2020-infuse.eu

Imagine a robot exploring Mars — it has cameras, laser scanners, and motion sensors, but each one only sees part of the picture. InFuse built the software that combines all those sensor feeds into one reliable understanding of the environment, like giving the robot a single, sharp pair of eyes instead of six blurry ones. The result is an open-source toolkit that any robotics team can plug into their system, whether the robot works in orbit, on a planet's surface, or in harsh conditions on Earth. Think of it as a universal translator that lets different sensors talk to each other and agree on what they're seeing.

By the numbers

consortium partners contributing to the CDFF

countries represented in the consortium (BE, DE, FR, UK)

total deliverables produced

SMEs in the consortium including the coordinator

The business problem

What needed solving

Robots operating in space or extreme environments rely on multiple sensors — cameras, laser scanners, inertial units — but each sensor produces its own separate data stream. Without proper fusion software, robots make poor navigation decisions, miss obstacles, or fail to build accurate maps. Building custom sensor fusion from scratch for each mission is expensive, slow, and error-prone.

The solution

What was built

The project built the Common Data Fusion Framework (CDFF) — an open-source software toolkit that combines data from multiple sensors into unified perception and navigation outputs for robots. The final package includes the complete software release plus an archive of testing datasets made available to the robotics community, totaling 17 deliverables.

Audience

Who needs this

Space robotics companies building autonomous rovers or orbital servicing vehiclesIndustrial robotics firms deploying autonomous systems in GPS-denied environmentsDefense contractors developing unmanned ground vehicles for reconnaissanceMining and offshore companies using robots for remote inspection in hazardous areasRobotics startups needing validated sensor fusion without building from zero

Business applications

Who can put this to work

Space Robotics & Satellite Services

SME

Target: Companies building autonomous spacecraft, orbital servicing robots, or planetary rovers

If you are a space robotics company struggling to integrate data from multiple sensors on autonomous vehicles — this project developed the Common Data Fusion Framework (CDFF), an open-source software package that combines perception and navigation data from diverse sensors into a unified picture. The final release was designed as a reusable baseline, cutting your development time for sensor integration on both orbital and planetary missions.

Industrial Automation & Inspection

mid-size

Target: Companies deploying autonomous robots for inspection, maintenance, or hazardous environments

If you are an automation company deploying robots in environments where GPS is unavailable — like mines, tunnels, or offshore platforms — this project built sensor fusion components for navigation and environment mapping that work without external positioning. The CDFF handles on-demand retrieval of maps and 3D models, which means your robots can build and update their own understanding of the workspace in real time.

Defense & Security Robotics

enterprise

Target: Companies developing unmanned ground or aerial vehicles for reconnaissance and search-and-rescue

If you are a defense or security technology firm dealing with unreliable sensor readings in unstructured terrain — this project delivered 17 components covering perception, navigation, and data product management for robotic systems. The open-source CDFF was tested against archived datasets and built to handle multiple robotic scenarios, giving your engineering team a validated starting point instead of building sensor fusion from scratch.

Frequently asked

Quick answers

What would it cost to adopt this technology?

The CDFF Final Package was released as open source, meaning the software itself is free to use. Your costs would be integration engineering — adapting the components to your specific sensors and robotic platform. With 6 consortium partners having already validated it, integration documentation should be available.

Can this scale to industrial production environments?

The CDFF was designed as a reusable baseline for the broader Space Robotics Cluster (SRC), meaning it was built for adoption across multiple missions and teams. The architecture supports both on-orbit and planetary scenarios, suggesting it handles diverse operational conditions. Scaling to terrestrial industrial use would require adaptation but the modular design supports it.

What is the IP and licensing situation?

The CDFF Final Package is described as an open-source release. This typically means you can use, modify, and build upon the software, though the specific open-source license terms should be verified with the coordinator, Space Applications Services NV in Belgium. The testing datasets were also made available to the robotics community.

How mature is this — can I use it today?

The project closed in January 2019 with a final open-source package and validated test datasets. Based on available project data, the CDFF was tested against real sensor datasets and designed for adoption by the wider space robotics community. It is a tested toolkit, not a turnkey product — you would need engineering effort to deploy it in your specific use case.

What sensors and data types does it support?

Based on the project objective, the CDFF covers perception and navigation sensor data processing, including environment mapping, object modeling, and science data. The EuroSciVoc tags confirm coverage of data processing, sensors, and robotics. Specific sensor types (LiDAR, cameras, IMUs) should be confirmed from the 17 deliverable reports.

Who built this and can they help with integration?

The consortium of 6 partners across 4 countries includes 2 industry players (both SMEs), 1 university, and 3 research organizations. The coordinator, Space Applications Services NV, is a Belgian SME with direct space industry experience. They would be the primary contact for technical support or integration consulting.

Consortium

Who built it

The InFuse consortium is a compact team of 6 partners across 4 countries (Belgium, Germany, France, UK), with a 33% industry ratio. The coordinator, Space Applications Services NV, is a Belgian SME — meaning the project was led by a company with commercial skin in the game, not just a university lab. With 2 industry partners, 3 research organizations, and 1 university, the mix leans toward R&D but the SME leadership suggests practical, market-oriented decision-making. For a business looking to adopt this technology, the coordinator is the natural first contact — they understand both the technical depth and commercial realities.

SPACE APPLICATIONS SERVICES NVCoordinator · BE
DEUTSCHES FORSCHUNGSZENTRUM FUR KUNSTLICHE INTELLIGENZ GMBHparticipant · DE
DEUTSCHES ZENTRUM FUR LUFT - UND RAUMFAHRT EVparticipant · DE
MAGELLIUM SASparticipant · FR
CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE CNRSparticipant · FR
UNIVERSITY OF STRATHCLYDEparticipant · UK

How to reach the team

Space Applications Services NV is a Belgian SME specializing in space applications — reachable through their corporate website or the project page

Order a report on this consortium See SPACE APPLICATIONS SERVICES NV profile →

Next steps

Talk to the team behind this work.

Want an introduction to the InFuse team for sensor fusion integration? SciTransfer can connect you directly with the right technical contact.

Order a report on this topic More in Digital & ICT