Adaptable Room Automation Using Multi-agent System

Building automation (BA) systems ensure comfort and safety of
the occupants while optimizing energy consumption. However, the
requirements of space comfort and the capabilities of the equipment which serve it often change during the lifetime of the building.
This is seen especially for systems like heating, ventilation, and
air conditioning (HVAC), lighting, etc. in rooms . The current approach of manually re-engineering the room automation system
to realign with the changes is time-consuming (and hence, higher
costs), and misaligned automation can result in energy loss. To
address this challenge, we show how goal-driven multi-agent systems which use procedural reasoning in their planning process
enable automation systems to recognize and dynamically readjust
to changing requirements and environment capabilities. The novelty of our approach lies in decomposing the available high-level
feature requirements of the rooms to goals for the agents while
providing it semantic description of the room environment . We
validated our approach by implementing an automation system for
a real-life office room and subjecting it to changes in requirements
and equipment capabilities