ESA GNC Conference Papers Repository

In the frame of preparation of future exploration missions requiring a high level of autonomy, the versatility of vision sensors is an asset clearly identified by the European Space Agency in the roadmaps. From February 2008 to September 2009, the study “Multipurpose Vision-based Navigation Sensor Architecture Definition” led by Astrium – called VisNav Architecture for short – has defined the building blocks for a modular approach of the vision-based navigation chain, based on common generic parts (detector, processor, board) and limited mission-specific elements (optics, algorithms). The first section of this paper reviews the exploration missions for which the VisNav navigation chain has been designed: this encompass planetary landing missions, rover missions, autonomous rendezvous missions and also interplanetary navigation missions. For each of them, requirements have been derived to set the sizing constraints for a multi-purpose sensor chain. In the second section, the modular approach of the VisNav architecture is detailed. This architecture is built to answer the technological challenge of sharing high level of autonomy with complex operations while taking into account minimization of industrial development costs. Common components with clear interfaces are designed, and the remaining specific parts requiring extended tailoring for each mission are detailed. For each buildingblock – optics, camera detector, processing electronics, communication links – a baseline design is identified. The last section discusses the capabilities offered by the VisNav sensor architecture in terms of navigation, beyond the state of the art established in past ESA studies. This is the natural continuation of the VisNav logic applied to the camera and its associated processing electronic to the on-board navigation software, which is currently under development by Astrium.