This thesis is an explorative study of spatial encounters between older adults and autonomous robots. We have collected data from a wide range of contexts to explore the domain; using field studies, observations, focus groups, and interviews. Our primary empirical context has been an institution where older adults live independent in separate apartments, but with common areas for social activities. A robot platform based on the Robot Operating System framework has been used to observe how older adults react when encountering a robot. We have found that the deployment of an autonomous robot is a pervasive alteration in the life of an older adult, and that it will require a substantial amount of facilitating tasks to be able to move autonomously. Thorough investigations and collaboration with participants for each particular context are required prior to deployment of the robot. A robot should communicate congruent information, utilizing multiple communication modalities. We have seen that spatial conflicts can emerge from encounters between robots and older adults, and propose five design implications to mitigate these conflicts. The most important proposals are to identify possible problem areas, and develop navigation strategies that not only focuses on efficiency, but has an explicit focus on avoidance of such conflicts and challenges.