Energies 2015,8 11013 While the research topics highlighted above provide new categorical options for future research,there are still several lingering gaps in knowledge relevant to the three approaches discussed in the previous sections.The following subsections discuss the issues and challenges of each approach separately 5.1.1.Monitoring Occupant-Specific Energy Consumption In regard to monitoring occupant-specific energy consumption in commercial buildings,load disaggregation among individual occupants is still a challenging issue.Although the literature has demonstrated a large variety of occupancy-sensing techniques,very little research has been conducted in the area of monitoring occupant-specific energy consumption.In fact,building management systems have been utilizing increasingly extensive sensor networks,but these networks often fail to correctly collect building occupancy data [155]and therefore do not effectively leverage total energy consumption data as a measurement of individual occupant's energy consumption.The fact that there are so few publications about approaches for monitoring occupant-specific energy use [156-159],gives evidence to the fact that less attention has been paid to this approach than to the other two main approaches (i.e.,simulation and improvement of occupants'energy consumption).However,the success of simulation and improvement approaches highly depends on detailed occupant-specific energy consumption. In fact,outputs of monitoring individual occupant's energy consumption can form the inputs for the second and third approaches. Monitoring occupant-specific energy consumption also provides researchers with the ability to quantitatively classify occupants into different energy-related groups based on their specific energy-use behaviors.Such classifications could help improve occupant-driven energy-conserving behaviors. Furthermore,the outcomes of occupant-specific energy use would provide researchers with an opportunity to present explicit feedback to individual occupants about their own individual energy actions and decisions.Future research is therefore recommended to propose models and techniques that would monitor the energy load of individual occupants. One option for future actions would be to extend the concept ofexisting non-intrusive load monitoring techniques.Such NILM techniques have been widely employed to disaggregate total energy consumption to identify specific loads and subsequently individual users.This concept would be helpful for developing related reliable methods for estimating the energy consumption of individual occupants.Chen and Ahn [13] indicated that Wi-Fi connection/disconnection events could be an effective indicator for occupancy energy load variation in commercial buildings.Developing such occupancy frameworks as well as occupancy-detection technologies could also be helpful in developing occupancy non-intrusive load monitoring techniques. Furthermore,Gulbians et al.[160]recently proposed a three-stage clustering algorithm as a new set of metrics that classifies commercial building occupants according to their energy-use efficiency, entropy,and intensity.This algorithm segments building occupants'energy consumption data in order to understand individual occupant's energy-use characteristics.Further developing the concept of such algorithms in order to estimate energy-use information of individual occupants is recommended.Energies 2015, 8 11013 While the research topics highlighted above provide new categorical options for future research, there are still several lingering gaps in knowledge relevant to the three approaches discussed in the previous sections. The following subsections discuss the issues and challenges of each approach separately. 5.1.1. Monitoring Occupant-Specific Energy Consumption In regard to monitoring occupant-specific energy consumption in commercial buildings, load disaggregation among individual occupants is still a challenging issue. Although the literature has demonstrated a large variety of occupancy-sensing techniques, very little research has been conducted in the area of monitoring occupant-specific energy consumption. In fact, building management systems have been utilizing increasingly extensive sensor networks, but these networks often fail to correctly collect building occupancy data [155] and therefore do not effectively leverage total energy consumption data as a measurement of individual occupant’s energy consumption. The fact that there are so few publications about approaches for monitoring occupant-specific energy use [156–159], gives evidence to the fact that less attention has been paid to this approach than to the other two main approaches (i.e., simulation and improvement of occupants’ energy consumption). However, the success of simulation and improvement approaches highly depends on detailed occupant-specific energy consumption. In fact, outputs of monitoring individual occupant’s energy consumption can form the inputs for the second and third approaches. Monitoring occupant-specific energy consumption also provides researchers with the ability to quantitatively classify occupants into different energy-related groups based on their specific energy-use behaviors. Such classifications could help improve occupant-driven energy-conserving behaviors. Furthermore, the outcomes of occupant-specific energy use would provide researchers with an opportunity to present explicit feedback to individual occupants about their own individual energy actions and decisions. Future research is therefore recommended to propose models and techniques that would monitor the energy load of individual occupants. One option for future actions would be to extend the concept of existing non-intrusive load monitoring techniques. Such NILM techniques have been widely employed to disaggregate total energy consumption to identify specific loads and subsequently individual users. This concept would be helpful for developing related reliable methods for estimating the energy consumption of individual occupants. Chen and Ahn [13] indicated that Wi-Fi connection/disconnection events could be an effective indicator for occupancy energy load variation in commercial buildings. Developing such occupancy frameworks as well as occupancy-detection technologies could also be helpful in developing occupancy non-intrusive load monitoring techniques. Furthermore, Gulbians et al. [160] recently proposed a three-stage clustering algorithm as a new set of metrics that classifies commercial building occupants according to their energy-use efficiency, entropy, and intensity. This algorithm segments building occupants’ energy consumption data in order to understand individual occupant’s energy-use characteristics. Further developing the concept of such algorithms in order to estimate energy-use information of individual occupants is recommended