Optimizing and improving energy performance of buildings while maintaining occupants' comfort are primary goals for building managers. In order to analyze a building's energy performance and make informed retrofit, maintenance, or operational decisions, decision-makers need access to credible real-time data illustrating how building systems are being used by its occupants at a floor and room level granularity. Traditionally, such data has been collected in buildings using wired or wireless systems by installing a dense array of sensors in every building location that needs monitoring. This is an effort and cost-prohibitive approach, especially in existing older buildings where instrumentation and integration with existing building systems is challenging. This paper introduces a novel concept of using autonomous mobile indoor robots for monitoring various occupant comfort and energy parameters inside an existing building, and discusses how the collected data can be utilized in various analyses. The research evaluates the hypothesis that a single multi-sensor fused robotic data mule that collects building energy systems performance and occupancy comfort data at sparse locations inside a building can provide decision-makers with a rich data set that is comparable in fidelity to data obtained from pre-installed and fixed sensor systems. In order to demonstrate the effectiveness of the proposed approach, an experiment was conducted using a tele-operated robot outfitted with thermal comfort data collection sensors and a localization camera in a multi-occupancy space within a large university building. The data collected by the mobile robot was statistically compared with data obtained from the building's pre-installed Building Automation System. Experimental results demonstrated the proposed method's promise and applicability in collecting dense actionable data in large spaces using only a sparse set of sensors mounted on mobile indoor robots.