机器人
感觉系统
计算机科学
人机交互
工程类
模拟
人工智能
心理学
神经科学
作者
Blanca Hernández Quintana,Konstantin Vikhorev,Antonio Adán
标识
DOI:10.1016/j.buildenv.2021.108194
摘要
The majority of the techniques in the building comfort monitoring state-of-the-art are based on local/manual measurements or on permanent sensor networks. These techniques entail imprecision and randomness (in the first case) and high-cost installations and a lack of flexibility to eventual changes in buildings (in the second). However, intelligent mobile platforms are becoming significantly more important as they perform data acquisition adapted to specific scenarios and schedules. In this paper, we present a robotic platform focused on performing the workplace occupant comfort-monitoring process. The soundness of our proposal compared to others lies on that it gathers most of the necessary properties of an effective monitoring platform: it collects a wider range of variables; it autonomously navigates in inhabited buildings managing occlusions and unexpected events; it conducts multiple monitoring sessions in one or several days; it provides comfort evaluations. Additionally, it can be very useful for energy engineers and construction professional as it provides valuable information in regard to comfort: it detects the best/worst results of the tested variables, locates discomfort in specific areas and moments, recognizes discomfort patterns and globally classifies zones into comfort classes. This robotic platform has been successfully tested in the interiors of buildings, providing significant and clear results in comfort terms (a case study is presented in this paper). However, some limitations and improvements should be addressed. Among other aspects, the computer-robot communication robustness for long distances and the procedure for detecting of small obstacles must be improved in the future. • ComfBot is an autonomous (not commanded) mobile platform, which carries out a complete (not partial) monitoring of the scene. • ComfBot collects temperature, humidity, pressure, lighting, CO2 levels, TVOC levels and noise levels. • The system can conduct several scheduled monitoring sessions during a day, which are previously programmed by the user. • It is a portable and low-cost platform, which enhances the applicability and affordability of the system. • It has powerful visualisation tools, which allows the user to carry out an efficient analysis of the data collected. • The system is able to detect areas that may not be comfortable in specific locations and at specific times.
科研通智能强力驱动
Strongly Powered by AbleSci AI