The older population is growing and society is facing an increase of people with cognitive disabilities, e.g. dementia. Social robots have the potential of providing company, lower stress levels and decrease loneliness. But there is only one robot that has been tested with controlled long-term repeated studies [Wada et al., 2005]. Because of this, there are no evidence that social robots have the same impact beyond the novelty effect, which is the tendency for performance to initially improve when new technology is instituted, and not because of any actual improvement. Therefore, I propose long-term pre-test post-test controlled group design studies with elderly people in care homes. But I also find it important to design and implement robots and other smart-home technique that brings a meaning to the user. My first project would therefore be to conduct a workshop design study with elderly people, both living at home and in elderly care homes, to assess what they want, think and need. Secondly, I plan to conduct a study at an elderly care home facility with the following scenario: A humanoid robot interview the residents about their life, with the purpose to measure the effects for human-robot interaction over time.