Gaia
After making a major contribution to the mission’s scientific foresight as early as 1995, and right up to Gaia’s selection as the cornerstone of ESA’s Science Program in 2000, the GEPI team worked on many aspects of Gaia’s preparation: simulation of the instrument and observations, design and definition of the desired scientific specifications for the spectroscopic instrument, prototyping of the on-board data processing software, processing of double and multiple star data. Since 2006, GEPI has played an important role in the Gaia Data Processing and Analysis Consortium (DPAC).

The Gaia space observatory, a cornerstone of the European Space Agency (ESA), is an ambitious project to study our galaxy. The satellite was launched from Kourou by a Soyuz-Fregat rocket on December 19, 2013 at 10:12 Paris time. Its primary vocation is to survey over a billion stars in the Milky Way and measure their positions, distances, motions and physical properties with unrivalled accuracy (50 to 100 times better than that of Hipparcos, the first astrometric satellite. By combining astrometric, photometric and spectroscopic data, Gaia will provide an unprecedented wealth of information about our Galaxy, enabling a detailed study of its three-dimensional structure, kinematics, origin and evolution, as well as all the types of stars that make it up. Gaia will also count and measure a very large number of brown dwarfs, extrasolar planets, asteroids (especially NEOs), supernovae and galaxies, making a major contribution to extragalactic distance scaling and fundamental physics.
The satellite
The Gaia project has been built on the experience of the Hipparcos mission: the same care taken to examine the scientific potential of the project beforehand, the same principle for the satellite (regular scanning of the sky) and the telescope (two sky fields separated by a wide angle combined in the focal plane), in order to obtain high-precision global astrometry and make optimum use of the observation time available. From the outset, it also addressed the problems encountered by Hipparcos: systematic on-board detection (for improved Galaxy sampling statistics), enhanced photometric performance (to determine the physical characteristics of observed objects in parallel with astrometric parameters) and, last but not least, an on-board spectroscopic instrument to directly obtain Doppler radial velocities (the third component of stellar space velocity) and detailed information on the atmospheres of the brightest stars.
Advances in technology (detection modes, computing power available on board, etc. ) and much larger mirrors than Hipparcos (two 1.45 m × 0.50 m rectangular mirrors, compared with Hipparcos’ 29 cm diameter mirror) have led to a spectacular leap in both expected precision (0.01 to 0.025 milliseconds of degree at magnitude 15, depending on the color of the stars, and again from 0.13 to 0.6 milliseconds at magnitude 20, compared with 0.2 milliseconds of arc for magnitude 5 stars or 1 millisecond at magnitude 9 for Hipparcos) and in the number of observable objects (one billion objects compared with 118,000).
The illustrations below show :
an "exploded" view of the satellite, showing its various components,
a diagram of the two telescopes, with their two sets of mirrors and the light path for one of the sky fields.

After being reflected by four successive mirrors (from M1 or M’1 to M4 or M’4), light rays from two fields of sky follow the same path, being reflected by mirrors M5 and M6 and finally arriving at the focal plane. Some of the light rays pass through the prisms and lenses of the two spectrophotometers and the RVS (Radial Velocity Spectrometer) spectrograph,
_

a zoom on part of the payload showing the path of the various light beams arriving on the focal plane and the position of the two spectrophotometers and the RVS spectrograph.