The EasyLiving project at Microsoft Research focuses on developing an architecture and technologies for intelligent environments that allow the dynamic aggregation of diverse input/output (I/O) devices into a single coherent user experience. The system includes components such as middleware for distributed computing, world modeling for location-based context, perception for collecting world state information, and service description for device control and user interface. This paper outlines current research in these areas and highlights common requirements for intelligent environments. The project aims to enable traditional PC-focused activities to move beyond a fixed desktop into the environment as a whole. Intelligent environments will include a variety of devices, both stationary and mobile, such as input devices like mice, keyboards, cameras, and active badge systems, and output devices like speakers, displays, and lighting. The system must understand the physical space from both sensory and control perspectives to provide a seamless user experience. An example scenario illustrates how the system works: Tom uses a PC to manage his music playlist and interacts with a wall screen to control room lighting. Sally enters the room, logs in, and uses the wall screen to view a document. The system must support these interactions across a distributed set of devices. The project requires research in middleware, geometric world modeling, perception, and service description. Middleware is essential for supporting traditional operating system functions across a networked, heterogeneous set of devices. The goal is to create an architecture that aggregates diverse devices into a coherent user experience.The EasyLiving project at Microsoft Research focuses on developing an architecture and technologies for intelligent environments that allow the dynamic aggregation of diverse input/output (I/O) devices into a single coherent user experience. The system includes components such as middleware for distributed computing, world modeling for location-based context, perception for collecting world state information, and service description for device control and user interface. This paper outlines current research in these areas and highlights common requirements for intelligent environments. The project aims to enable traditional PC-focused activities to move beyond a fixed desktop into the environment as a whole. Intelligent environments will include a variety of devices, both stationary and mobile, such as input devices like mice, keyboards, cameras, and active badge systems, and output devices like speakers, displays, and lighting. The system must understand the physical space from both sensory and control perspectives to provide a seamless user experience. An example scenario illustrates how the system works: Tom uses a PC to manage his music playlist and interacts with a wall screen to control room lighting. Sally enters the room, logs in, and uses the wall screen to view a document. The system must support these interactions across a distributed set of devices. The project requires research in middleware, geometric world modeling, perception, and service description. Middleware is essential for supporting traditional operating system functions across a networked, heterogeneous set of devices. The goal is to create an architecture that aggregates diverse devices into a coherent user experience.