PaperlayerView exact revision on arXiv

Source-linked AI summary

An Internet of Things Framework for Smart Energy in Buildings: Designs, Prototype, and Experiments

Jianli Pan, Raj Jain, Subharthi Paul, Tam Vu, Abusayeed Saifullah, Mo Sha

arXiv:1603.08025v1cs.NIcs.CY

TL;DR

Building energy efficiency matters because buildings are major energy and environmental contributors, yet green-by-design buildings may operate inefficiently under centralized, static controls. This paper analyzes year-long measurements from a green office building, proposes smartphone- and cloud-enabled location-based control, and evaluates a prototype whose measured consumption approaches the frugal mode while supporting multi-scale energy proportionality.

  • Problem

    Buildings have major environmental and energy impacts, while centralized and static controls can leave green buildings inefficient during actual operation.

  • Method

    The paper combines year-long energy monitoring and analysis with a location-based IoT control framework using smartphones, distributed control, cloud computing, and an experimental prototype.

  • Results

    The prototype’s real recorded energy consumption after applying the location-based idea is very close to the frugal mode’s energy consumption.

  • Takeaways & Limitations

    The framework supports multi-scale energy proportionality across building, user, and organizational levels and has potential economic and social sustainability benefits.

Abstract

from arXiv · show

Smart energy in buildings is an important research area of Internet of Things (IoT). Buildings as important parts of the smart grids, their energy efficiency is vital for the environment and global sustainability. Using a LEED-gold-certificated green office building, we built a unique IoT experimental testbed for our energy efficiency and building intelligence research. We first monitor and collect one-year-long building energy usage data and then systematically evaluate and analyze them. The results show that due to the centralized and static building controls, the actual running of green buildings may not be energy efficient even though they may be "green" by design. Inspired by "energy proportional computing" in modern computers, we propose a IoT framework with smart location-based automated and networked energy control, which uses smartphone platform and cloud computing technologies to enable multi-scale energy proportionality including building-, user-, and organizational-level energy proportionality. We further build a proof-of-concept IoT network and control system prototype and carried out real-world experiments which demonstrate the effectiveness of the proposed solution. We envision that the broad application of the proposed solution has not only led to significant economic benefits in term of energy saving, improving home/office network intelligence, but also bought in a huge social implication in terms of global sustainability.

I. INTRODUCTION

Buildings have substantial environmental and energy impacts, while centralized and fixed controls can leave even green buildings inefficient in operation. The paper combines measured-data analysis, a location-based IoT control framework, and a prototype to pursue multi-scale energy proportionality.

  • Motivation: Buildings account for major shares of U.S. carbon dioxide emissions, electricity consumption, energy usage, water consumption, and non-industrial waste.The supplied survey reports approximately 38%, 71%, 39%, 12%, and 40%, respectively.
  • Motivation: Systematic monitoring and modeling are needed because buildings are complex, conventional controls lack intelligence, and energy consumption depends on multiple factors.The framework considers monitoring at granularities from whole buildings to occupants, followed by modeling and practical strategy adjustments.
  • Findings: One year of building-energy traces showed that centralized and fixed-pattern control can make green buildings inefficient despite green-by-design features.The paper uses these findings to motivate more adaptive energy control.
  • Framework: The proposed IoT framework combines smartphone location services, distributed control, cloud computing, and policies from individuals and organizations.These components support active occupant participation and integrated energy-saving decisions.
  • Contributions: The journal version integrates problem identification, solution design, prototype development, and experimental validation into a complete IoT framework.The paper positions this complete process as an expansion of earlier separate contributions.
  • Framework: A prototype demonstrates real-time, location-based automated energy-policy control across multiple buildings, replacing centralized static modes with distributed dynamic control.The design targets consumer-side smart grids and supports building-, user-, and organizational-level energy proportionality.

II. ENERGY EFFICIENCY EVALUATION AND ANALYSIS

The study develops a real-building energy-monitoring testbed and analyzes measured data to identify consumption patterns and their relationships with environmental conditions and occupancy. It uses these findings to support subsequent energy-control design.

  • Energy Monitoring Testbed and Justification: The testbed is a typical U.S. office building selected after consulting maintenance experts and comparing common building structures.The study also investigated a Net-Zero Energy Building as an extreme comparison case.
  • Energy Monitoring Testbed and Justification: The LEED Gold building logs overall resource usage through networked meters every 30 minutes, with some measurements recorded every 15 minutes.The monitored facility includes typical subsystems such as HVAC.
  • Data and Variables: The analyzed measurements include total electricity, heating and cooling energy, and indoor and outdoor temperature and humidity.HVAC consumption is represented by heating and cooling, while total electricity covers a broader range of building loads.
  • Methodology: The evaluation examines how energy consumption relates to environmental factors and occupancy rate.The stated goal is to identify consumption patterns and quantify relevant influences.
  • Methodology: The methodology combines short- and long-period correlation analyses, multiple time granularities, and Multiple Polynomial Regression and Multiple Linear Regression models.Hourly data are grouped into weekly and monthly granularities to reveal longer-term correlation differences.

C. Detailed Evaluation and Analysis

The evaluation finds weak links between building energy use and weather or occupancy, alongside substantial fixed consumption outside expected usage periods. Regression and seasonal analyses indicate that energy subsystems respond differently, motivating separate control adjustments.

  • Environmental Impacts Analysis: Electrical energy shows mostly weak correlations with temperature and humidity, generally below 0.5, while heating also lacks strong correlation with temperature.The study reports that overall weather correlations are small and Fig. 2 shows no very strong heating-temperature relationship.
  • Environmental Impacts Analysis: Cooling consumed 16.9 billion BTU versus 8.1 billion BTU for heating during the analyzed period, with summer cooling significantly higher than other months.Daily electrical consumption fluctuated regularly, but seasonality was less obvious than for heating and cooling.
  • Regression Modeling and Analysis: Temperature and humidity explain 19.02% of electrical-energy variation but 98.84% of cooling-energy variation under MPR regression.The contrast supports analyzing and tuning building energy subsystems separately because environmental factors affect them differently.
  • Occupancy Impact Analysis: Office-hour electrical consumption is about 15% higher than after-hours and weekend consumption, yet off-period usage remains higher than expected.The authors interpret this as evidence that fixed provisioning supplies extra capacity and wastes energy, especially after hours.
  • Occupancy Impact Analysis: After-hours heating is about 6% higher than weekend use and 19% higher than business-hours use, making business-hours heating the lowest.The authors suggest occupancy-related body heat may reduce external heating demand during business hours, while overall occupancy has very low impact on energy consumption.
  • Summary of Findings: The building’s centralized, fixed operating patterns produce poor operational efficiency despite green design, with energy consumption weakly connected to actual occupancy.The analysis motivates location-aware, automated control strategies that adjust building energy use more proportionally to conditions and usage.

III. SMART LOCATION-BASED AUTOMATED ENERGY CONTROL FRAMEWORK

The paper presents a smart location-based automated energy control IoT framework. The framework is designed as a central contribution of the work.

  • The framework uses location information to automate energy-control decisions.
  • It is presented as an IoT framework for building energy management.
  • The framework section builds on the authors’ earlier contributions.

A. Overall Structure

The proposed design combines interacting components into an occupant-oriented networked system. Its components cover monitoring, location awareness, policy management, cloud services, and energy modeling.

  • The framework is organized as an occupant-oriented networked system whose components interact.
  • The design is depicted as a complete framework assembled from multiple interacting aspects.
  • Its key components include mobile distributed monitoring, remote control, smartphone location services, policy strategies, cloud computing, and energy modeling.

B. Smart Mobile IoT Devices as Remote Controls

Smartphones serve as remotely accessible interfaces for building energy monitoring and control. Their connectivity and authentication support occupant participation and flexible policy changes.

  • Smartphones enable occupants to monitor, control, and manage energy systems remotely.
  • Authenticated occupants can modify energy-saving policies online through building policy servers.
  • The design supports dynamic policy changes and greater flexibility for occupants.

C. Multi-source Energy-saving Policies Hierarchy

The framework coordinates location-aware devices with hierarchical, multi-source energy policies across buildings and organizational levels. User movement can trigger policy-control actions between buildings.

  • Multi-source policies: Different organizational units and household members may contribute policies and requirements to energy control.
  • Policy hierarchy: Building control can use a tree-like hierarchy of policy servers covering different levels.
  • Policy hierarchy: Control regions are modeled as realms managed by realm managers, with conflicts resolved when policies span multiple realms.
  • Smart location awareness: Smartphone location services support automatic control based on a user’s position and direction of movement.
  • Cross-building coordination: Movement between home and office threshold ranges can send a server message that triggers policy control.

E. Cloud Computing and Storage

The framework uses cloud storage and computation alongside building-side servers, smartphones, and location-aware controllers. A prototype demonstrates automated appliance control across a user’s home and office.

  • Cloud Computing and Storage: Cloud services store and retrieve logged building-energy data and perform computation-intensive modeling and analysis.The communication layer is designed to provide configurability and reliability.
  • Location-Based Control: Location changes trigger energy-policy updates that turn associated home and office appliances on or off.The system is intended to support real-time user energy policies and consumption proportional to actual usage.
  • Prototype Scope: The proof-of-concept implements one user controlling devices in one home building and one office building.The authors describe this as a small-scale system for comparing savings against operation without the new design.
  • Prototype Networking: The prototype connects mobile location sensors, web servers, controllers, appliances, and WiFi routers across two buildings.Port mapping enables access to building-side servers from outside network address translation boundaries.

B. Experiments and Results

The experiments combine baseline appliance measurements, usage-mode estimates, and a 24-hour location trace to evaluate dynamic control in home and office settings. The prototype’s recorded consumption closely approaches the estimated frugal mode.

  • Experimental Design: The evaluation compares luxury, moderate, and frugal daily energy-use estimates with real consumption after location-based control.Home and office appliance baselines are documented separately.
  • Experimental Design: A 24-hour location trace drives dynamic appliance control in both home and office environments.The experiment applies policy changes according to the user’s tracked location.
  • Results: 14 hours at home produced 5.285 kWh of recorded consumption, including 2.7 kWh for lighting and 2.22 kWh for refrigeration.The home period included sleep, lunch and rest, and working at home.
  • Results: 6 hours in the office produced 2.26 kWh, including 1.15 kWh for lighting, 0.96 kWh for desktop use, and 0.15 kWh for laptop use.The office appliances were controlled according to the user’s location and activity.
  • Results: The prototype’s real energy consumption was very close to the frugal mode’s estimated consumption.The comparison is presented as evidence of energy savings from the location-based idea.

C. System Implementation and Integration Challenges

Implementation required integrating heterogeneous mobile, networking, server, controller, and appliance components while handling GPS inaccuracies. Threshold tuning was used to reduce false alarms before the experiments.

  • Control and Integration: The system had to support platform-independent remote control through general PC platforms rather than dedicated mobile platforms.This requirement motivated the search for suitable control devices and methods.
  • Control and Integration: The prototype integrates GPS-enabled mobile devices, data-upload modules, web servers, controllers, appliances, and routers with port mapping.Python implements distance calculation and threshold comparison before policy changes are triggered.
  • GPS Location Inaccuracy: GPS coordinates may vary by 20 meters even when the mobile user remains stationary.The authors tuned the threshold distance and conducted experiments to minimize false alarms or false positives.

D. Multi-scale Energy Proportionality

The experiments illustrate user-scale energy proportionality, while the framework extends the same idea to organizational and building levels. The authors associate this with potential economic, operational, and sustainability benefits.

  • User-Level Energy Proportionality: The user’s energy consumption becomes approximately proportional to actual usage through networked, location-based control.This is presented as user-level energy proportionality demonstrated by the prototype experiments.
  • Multi-Scale Proportionality: Applying the approach across organizational structures can enable organization-level and building-level energy proportionality.The framework describes multi-scale policies operating across these levels.
  • Implications: Multi-scale energy proportionality is associated with avoiding energy waste and saving costs for users and organizations.The networking and computing technologies are also described as supporting more intelligent, efficient, and automated building operation.
  • Implications: The proposed system involves users and organizations in energy-saving efforts and may encourage participation in climate and sustainability issues.The paper presents this as a potential educational and social benefit.

V. RELATED WORK

Related work spans building energy simulation, climate-effect modeling, wireless sensing, and mobile/cloud computing. The paper positions real-network monitoring and broad IoT integration as complements to these approaches for building energy control.

  • Building energy simulation: Building energy simulation software uses building parameters to estimate energy usage, whereas real-network monitoring analyzes actual consumption data.The paper states that monitoring and analyzing real energy data can be more effective than simulation alone.
  • Climate effect models: Climate-effect research studies how weather, heat transfer, building structures, and solar effects influence building energy efficiency.The cited work is described as largely theoretical.
  • Wireless sensing: Wireless Sensor Networks have been applied to building subsystems including lighting, thermostats, energy monitoring, activity logging, and HVAC scheduling.These systems can complement other sensing and metering technologies in building environments.
  • Mobile and cloud technologies: Smartphone platforms provide sensors such as GPS for versatile applications, while wide-scale building energy auditing and control remain limited.Cloud computing is also identified as relevant to the paper’s research.
  • Paper positioning: The paper combines energy-data analysis, location-based automated control design, and an experimental IoT prototype within a complete three-step research program.Its intended scope includes multi-scale energy proportionality and intelligent home or office spaces.
Loading 1603.08025v1…