Infusing Data Fusion in Space Robotics

Space robots explore celestial bodies and help keep man-made satellites orbiting the Earth in tip-top shape. EU-funded scientists have significantly enhanced their data-processing capabilities to help them do their jobs even better.

Space robots play an important role in advancing understanding of our cosmos and providing services for people on Earth. Planetary spacecraft (robots) such as orbiters, landers, and rovers conduct fundamental research of deep space bodies like the Moon, Mars, and even asteroids. Orbital space robots are machines designed to maintain or service spacecraft that are in Earth orbit, such as communications satellites or the International Space Station.

What all these robots have in common is the acquisition, processing, and utilisation of vast quantities of real-time data to do their jobs well. To help space robots make sense of a wealth of environmental data perceived through a variety of sensors, the EU-funded InFuse project created a common data fusion framework (CDFF).

Supporting the space robotics community

According to project coordinator Jeremi Gancet, “InFuse consists of a set of software algorithms and tools allowing assembly of elementary data fusion functions into larger modules addressing specific data fusion needs.” There are over 15 data fusion algorithms integrating information from more than 40 data-fusion sensor nodes.

The InFuse methods and tools will make it much easier to develop software for space and to evaluate it early in prototyping, streamlining the process in terms of costs and time. To maximise its uptake in both the space and non-space sectors, the consortium has made the CDFF available as open-source software. Non-space terrestrial applications include inspection robots (aerial, underwater, or ground-based), autonomous cars, and agile robotic manufacturing.

Fusing more than data

In addition to making the CDFF freely accessible to the space industry and beyond, InFuse is uniting the Horizon 2020 Space Strategic Research Clusters (SRC) on Space Robotics. As one of the six initial operational grants (OGs) in the SRC aimed at technological challenges, InFuse’s CDFF is being incorporated into the software architecture of all the second-wave OGs’ robotics platforms.

Coordinating research in large consortiums is challenging enough. Synchronising the work of all the OGs within the SRC posed additional challenges. However, with frequent discussions and interactions, the SRC created a stimulating working environment that achieved excellent results.

For example, 3D models of the environment and estimates of the robot position for localisation are helping sister project ERGO plan and verify robot paths for goal-oriented autonomous exploration. InFuse is also supporting the assimilation and processing of data produced by sensors in the I3DS project. The entire software platform will be integrated into the ESROCOS space robotics control and operating system developed specifically for space and other critical applications.

Testing and validation of the CDFF in planetary rover prototypes at a Mars-like environment in Morocco in the autumn of 2018 was a milestone. It provided an opportunity to collect realistic data and confirmed that the project’s mission was accomplished.

Although the project has officially ended, InFuse plans to continue evolution of its CDFF, further enhancing existing software and adding new data-fusion algorithms. In the case of InFuse, the sky is not the limit.

Reference source: Infusing Data Fusion in Space Robotics