Smart spaces provide enhanced user experience through sensing and adaptation to changing context. Hence, they allow distributed applications to show intelligent, autonomous and interactive behavior. Two important research topics within this field are machine learning and human-system interaction. This paper introduces a pilot smart space implementation on top of a Wireless Sensor and Actuator Network (WSAN) infrastructure that is specifically intended to investigate these two aspects. We investigate a scenario built around an office breakout area context. We describe the infrastructure setup, the novel user interfaces, the network architecture and the message exchange policies of the pilot implementation. We employ the opportunities of modern lighting systems in the proposed implementation.