One may simply be amazed in front of the diversity and complexity of life. Yet, and maybe even more bewildering, living systems all share common hallmarks: compartmentalization, growth and division, information processing, energy transduction and adaptability. In particular, the mobility plays a crucial role in the competitiveness between different species. Physics at microscales is different from the one we are used to at our macroscopic scale. This is why, micro-swimmers have developed specific strategies to induce motion. The understanding of such strategies is crucial at the fundamental level to apprehend the behavior of biological micro-swimmer, but also to achieve artificial locomotion in a surrounding fluid at the micron-scale, in order to perform a multitude of tasks in technical and medical applications (transport, mixing), which has become a central goal of nanoscience. In this context, biological and artificial micro-swimmers have been intensively studied, and we place our study in the framework of swimming in a realistic and complex environment, in the case where external factors (confinement, external force, other swimmers) may influence the swimming properties. In this work, using microfluidic, we create, put into complex situation and observe a model swimmer: a pure water swimming droplet in an outer oil-micelle solution. It was shown that the droplet motion emerges from the nonlinear coupling of hydrodynamics and advection-diffusion of micelles filled with water. We first study the effect of confinement on such geometries using confocal PIV in 3D. The presence of one wall breaks the natural axisymmetry of the flow field. We propose a simplified analytical formulation taking into account the presence of the wall and the effect of buoyancy. This model accounts for the far field hydrodynamic of the droplet close to a wall that differs from the no-wall case. We then look at more confined geometries using glass capillary microfluidic. The velocity of the droplet decreases with increasing confinement; but surprisingly; it saturates at a non-zero value for droplets bigger than the channel height: even very long droplets swim. In more complex geometries, such as stretched capillaries; the droplet elongates while swimming, and amazingly, may undergo successive spontaneous splitting events for high enough confinement. We show that this behavior comes from a saturation in the swollen micelles concentration along the droplet length. External force - such as gravity – also influence the droplet behavior. In 2D, by observing a swimming droplet on an inclined plane, we show that gravity orients the droplet, and that under strong gravity, the droplet’s velocity is more than the simple additivity of the gravity and activity. This is discussed in the light of a theoretical study of the instability mechanism under an external force. The droplet in 1D exhibit a similar behavior, but is also able to swim against gravity. Finally, we investigate their collective dynamics in a 1D micro-fluidic channel. We observe experimentally a rich phenomenology: neighboring droplets align and form large trains. Exanimating the interactions between two "colliding" droplets shows that alignment rises from the interplay between velocity fluctuations and the absence of Galilean invariance. Taking these observations as the basis for a minimalistic 1D model of active particles and combining analytical and numerical arguments, we show that the system exhibits a transition to collective motion. Altogether, the swimming droplet shares numerous similarities with living system: compartmentalization (a droplet), division (under confinement), energy transduction (by thermodynamic relaxation) and adaptability (through the swimming). Beyond the simple understanding of our peculiar system, these studies give insight on various phenomena at the interface of hydrodynamics, physico-chemical engineering and active matter.