|
40% of the nations energy is used for heating, cooling, ventilating, and lighting buildings. Lighting and heating/cooling requirements are directly interrelated. In commercial buildings all lighting energy eventually becomes heat energy and must therefore be cooled away, adding additional energy requirements. The deefficacy of lighting systems is about two to six percent when measured from the powerplant (powerplant efficiency 25%; transmission loss to buildings 15%; conversion of electricity into light efficiency 25%; light provided for unoccupied areas and regardless of daylight availability adds further penalties). About 35% of all electric energy is used in office buildings and the majority of the electricity is used for lighting. Health, well-being, motivation and productivity of office workers (over half of the U.S. workforce of 120 Million people) depend significantly on the appropriate quantity and quality of lighting, including daylighting systems. Glare (reflected or indirect), illuminance and contrast values are critical.
OWL (Object-Oriented Workplace Laboratory) is an attempt to deal with these problems. OWL users have a high degree of control over their work space at the Intelligent Workplace. From their user interface, they may request changes in the temperature, light level, and other aspects of the environment. These changes are then effected through adjustments in exterior louvres, internal lighting systems, and the HVAC system. Through a network of environmental sensors located in each work area, an energy-efficient adjustment of building climate systems can be formulated and executed.To schedule workplace resources, a reservation system (aware of the day-to-day geometry of the Intelligent Workplace) assists users in requesting conference rooms and other shared resources.
The OWL system offers major advantages over standard lighting systems currently employed in commercial office buildings. The benefits are substantial energy savings, improved functionality and user satisfaction through the use of internet and componentware technologies.
Buildings offer major challenges because the human/machine interface is very complex (many users with different tastes and preferences, many parts/components/systems, many functional changes ). Therefore systems that allow for individual control, combined with embedded intelligence to guard against energy and environmental waste must be develeoped. The OWL system directly addresses these issues.
The OWL software architecture is designed as a framework that can be reused for other application domains that be modeled as a space of sensor objects, where we define a sensor as an object with several attributes whose values are read, visualized and controlled by remote method invocation. We believe the framework can be applied to other application domains with sensor spaces such as airplanes, cars, trucks, ships...not just buildings.
Based on a software architecture that combines Web (Java) and Object technology (CORBA), OWL offers opportunities for remote diagnostics, remote facility management, remote control and remote visualization of lighting systems in the workplace. OWL's software architecture provides flexibility as well: As the operational needs of building occupants change, the interior modular walls may be rearranged to provide suitable work environments. These changes may be simulated in a CAD package for analysis of thermal efficiency implications, then visualized in 3D using the Liquid Reality toolkit.
CORBA allowed us a unique degree of collaboration amongst the OWL subsystems and hardware resources running on a variety of architectures and operating systems. The first demonstration of OWL held on December 5, 1996 employed the Janus speech recognition system written in tcl/tk and a graphical User Interface written in Java running under Solaris on a Sun SPARCstation. This interface obtained building geometry and environmental information via the object request broker from an object-oriented database running on a remote HP workstation. The Database Subsystem (RAD and API) was accessed by several other OWL subsystems each of them accessing existing software products running on different plaforms and often written in a different language: the Facility Management subsystem providing access to Reservation and Site Editors (Site Editor object model and RAD) written in Java running on a Windows NT Pentium PC; the thermal Simulation subsystem (RAD) written in tcl/tk, employing a CAD tool - Microstation - running on an NT system, but whose computation is performed by a remote Unix server; the Control subsystem written in Java and interacting directly with the building sensors, louvres, and other Intelligent Workplace hardware through an NT workstation; and the Visualization subsystem building VRML representations of CAD models stored in the database running on a Solaris SPARCstation.
Communication between the OWL subsystems is provided by a powerful event service called Notification. The event service is based on a push architecture, helping to coordinate computation, data storage and physical manipulation of the Intelligent Workplace hardware.
The development of OWL followed an incremental and iterative lifecycle using a hybrid design methodology based on Jacobsen's use cases and Rumbaugh's OMT. (Note that we did not use the Unified Method Notation proposed by Booch, Rumbaugh and Jacobsen). The methodology was taught in lectures given in parallel while the system was being developed. Starting with a problem statement, the first iteration OWL (Object-Oriented Workplace laboratory) was developed from August to December 1996 for the Intelligent Workplace, a testbed of advanced building products and systems at the Center for Building Performance and Diagnostics at Carnegie Mellon University. The first iteration of the OWL system produced several work products such as a software project management plan, a requirements analysis document, system design document, object design document and test manuals. Two preliminary prototypes of the User interface and the visualization subsystem were developed. The system prototype was delivered during the client acceptance test on Dec 5, 1996. The status of the system and the documentation is described in the status page.
The development of the first iteration took 3 months involving more than 80 people (People and Clients). The communication was based on face-to-face group meetings, project-wide reviews and distributed communication via NetNews . The second iteration is the topic of the Advanced Software Engineering Course.
Design patterns were used to interface to existing systems built by other projects at Carnegie Mellon University (NODEM, Janus) and existing software products available from commercial vendors (Microstation). Java (Java I and Java II slides) was used as implementation language whenever possible. Combined with the fact that most new code written for the project was implemented in Java, the CORBA/web approach proved to be the best solution for an extremely flexible, platform-independent system.
|
|