The RUBICON project is a three-year, European Union-funded project to minimize the need for programming and human supervision of robotic ecologies. The RUBICON ecology is able to adapt to changes in its application requirements and environment. The high-level RUBICON architecture includes control, cognitive, and learning layers to allow the system to execute tasks, sense its environment, and learn from feedback and training. An ambient assisted living scenario was implemented in a homelab at TECNALIA using sensors, actuators and a Turtlebot to capture data and interact with users. The results achieved adaptive and reconfigurable distributed learning, and experimental assessment of localization and adapting to environmental changes.