Sensor systems in the built environment

Sensor systems can play useful roles towards achieving several different objectives in the built environment. These objectives include energy conservation, improved indoor air quality, and, more recently, protecting buildings from intentional and harmful releases of chemical or biological contaminants. Along the lines of building “security”, fire and earthquakes also pose physical threats and systems are developed and deployed in response to these concerns. To promote energy efficient, healthy and secure building operation, real-time monitoring systems can be extremely helpful and in some cases may be essential. For example, it would be difficult to optimize thermal-mechanical system performance according to building thermal loads without sensing and control systems. Similarly, sensor systems can be invaluable by identifying localized security threats such

as fire and chemical releases. This section reviews sensor systems that have been employed for achieving these objectives and provides the broader context in which sensor systems play a role in the built environment.

Increasing energy efficiency has been a strong motivation in the development of indoor sensor and monitoring systems. Heating, ventilating and cooling (HVAC) systems, along with lighting systems, account for a large percentage of commercial building energy consumption. The advent and widespread deployment of digital technology initiated a transformation in the 1980s in how HVAC systems are controlled: pneumatic control systems were replaced with digital control systems.

With digital controls, an energy management and control system (EMCS) acts as a supervisory controller. In addition to controlling thermal equipment, the EMCS also can collect and monitor the performance of building energy systems. However, research has shown that “typical” practice does not fully utilize the features of the EMCS (Piette et al., 2001), and, thus, the hopes of reduced energy consumption with digital technology have yet to be substantially realized.

The energy efficiency research community has been addressing the problem of poor control and operation of building systems and underutilization of available technology. Examples of research efforts include the deployment of field-scale monitoring systems (Piette et al., 2001) and development of automated diagnostics of HVAC systems (Katipamula and Brambley, 2005a and 2005b).

The literature in the automated diagnostics area, sometimes referred to as fault detection and diagnostics, is extensive. The concept is to use sensor information — sometimes available through the HVAC control systems themselves — and to determine

if there is faulty operation. Faulty operation can lead to increased energy consumption, additional maintenance costs, and decreased occupant comfort. Katipamula and Brambley (2005a; 2005b) provide a comprehensive review of the advances in this field.

Basic monitoring efforts have illustrated energy savings opportunities through simple improved operation. Piette et al. (2001) describe the use of a web-based tool that operates in parallel with the EMCS and diagnoses problems in building HVAC systems.

More recently, efforts have been made to use wireless sensor networks for improving building control, both in lighting and in HVAC systems. Kintner-Meyer and Conant (2004) describe the integration of wireless sensors into HVAC controls. Granderson et al. (2004) describe the integration of wireless sensor networks for advanced lighting control based on incorporating user preferences through a Bayesian decision tool.

Among the research efforts described in this subsection, methods employed in fault detection and diagnostics most closely resemble the research presented in this dissertation. One approach to automated diagnostics uses sensor information to estimate unobserved system parameters and then uses these parameters to indicate the presence of faulty operation. The problem is often formulated as an inverse problem, since the parameters of interest are not typically directly measured.

Compared to the time scales of interest in this dissertation, the time scales encountered in fault detection and diagnostics are usually longer. For example, the fouling of a heat exchanger is likely to occur over time scales of months to years. As a result, real-time computational methods may not be required. Methods range from physical and analytical modeling, to those using artificial intelligence or statistical techniques. Qualitative methods, such as rule-based systems, have also been used

(Katipamula and Brambley, 2005a and 2005b). So far, it does not appear that Bayesian methods have been applied in the HVAC fault detection and diagnostics efforts.

In addition to the energy-related research, there are extensive literatures on intelligent buildings and on fire protection sensor systems. Jablonski et al. (2003) provide an overview of issues in designing an intelligent building. They discuss how, in an intelligent building, a unifying centralized information system controls all functions of the building, including fire protection, thermal condition, IT systems, and office automation. The realization of a complete intelligent building requires cooperation among a diverse and disparate range of disciplines and designers, which is rare within the building industry, particularly in the US.

Sensor systems for fire protection have many parallels with sensor systems for detecting high-risk contaminants. Liu and Kim (2003) provide an excellent summary of recent advances in automated systems used for fire protection, many of which are enabled by advances in sensor technology and by the use of artificial intelligence techniques for sensor information processing.

In summary, there are many objectives that have motivated or should motivate the development of sensor systems and the adoption of information technology in the built environment. The harnessing of information technology for increasing energy efficiency has not reached its full potential. In part, progress has been hampered by the lack of sufficient economic incentives, and because inefficient thermal and lighting systems are not life-threatening. In contrast to energy systems, the failure of security systems, such as fire protection or chemical detection systems, poses acute risks. It is possible that overall improvement in achieving building security objectives may encourage an

improvement in performance of other systems, thereby resulting in better performing buildings overall.