Alaa Alhamoud

Extracting Human Behavior Patterns from Appliance-level Power Consumption Data

In order to provide useful energy saving recommendations, energy management systems need a deep insight in the context of energy consumption. Getting those insights is rather difficult. Either exhaustive user questionnaires or the installation of hundreds of sensors are required in order to acquire this data. Measuring the energy consumption of a household is however required in order to find and realize saving potentials. Thus, we show how to gain insights in the context of energy consumption directly from the energy consumption profile. Our proposed methods are capable of determining the user’s current activity with an accuracy up to 98% as well as the user’s current place in a house with an accuracy up to 97%. Furthermore, our solution is capable of detecting anomalies in the energy consumption behavior. All this is mainly achieved with the energy consumption profile.

SMARTENERGY.KOM: An intelligent system for energy saving in smart home

Over the last twenty years, energy conservation has always been of great importance to individuals, societies and decision makers around the globe. As a result, IT researchers have shown a great interest in providing efficient, reliable and easy-to-use IT services which help users saving energy at home by making use of the current advances in Information and Communications Technology (ICT). Driven by the aforementioned motivation, we developed SMARTENERGY.KOM, our framework for realizing energy efficient smart homes based on wireless sensor networks and human activity detection. Our work is based on the idea that most of the user activities at home are related to a set of electrical appliances which are necessary to perform these activities. Therefore, we show how it is possible to detect the users current activity by monitoring his fine-grained appliance-level energy consumption. This relation between activities and electrical appliances makes it possible to detect appliances which could be wasting energy at home. Our framework is organized in two components. On one hand, the activity detection framework which is responsible for detecting the users current activity based on his energy consumption. On the other hand, the EnergyAdvisor framework which utilizes the activity detection for the purpose of recognizing the appliances which are wasting energy at home and informing the user about optimization potential.

Presence detection, identification and tracking in smart homes utilizing bluetooth enabled smartphones

Advances in ubiquitous computing over the last decade have allowed us to inch closer to the realization of true smart homes. Many sensors are already embedded in our living environments which can monitor several environmental parameters such as temperature, humidity, brightness and appliance-level power consumption. However, in order to achieve the primary goal of the smart home, we should be able to detect, identify, and localize the entities inside it. Therefore, the user detection, identification and localization problems represent a crucial facet of the challenges introduced by the smart home problem. Our approach towards solving these challenges entailed the usage of Bluetooth technology for user identification and tracking, alongside a Wireless Local Area Network setup to collate the sensor data at a centralized server such as a home gateway which subsequently processed and stored the entries. Moreover, we have studied the efficacy of various pattern recognition algorithms for real time processing and decision modeling on the received data. We have hence demonstrated our solution represents a non-intrusive, inexpensive and energy-conserving methodology to solve an essential part of the smart home problem by integrating already existent devices and infrastructure in an innocuous manner to obtain good results with minimum overhead.

Empirical investigation of the effect of the door's state on received signal strength in indoor environments at 2.4 GHz

Due to the wide deployment of indoor wireless local area networks (WLANs), the indoor planning became a research of interest for IT as well as networking researchers. As a result of this wide deployment, many IT applications and services started relying on the ready implemented WLAN infrastructure. Therefore, there is a need for reliable propagation models which are able to predict the WLAN signal strength in indoor environments before starting the real world deployment which leads to an efficient and cost aware deployment process. In this paper we develop an empirical propagation model which focuses mainly on the effect of the door state on the propagated WLAN signal in indoor environments. The measurements were compared to other simulated results in literature. A new empirical parameter based on empirical measurements was introduced for a better estimation of the received signal strength (RSS).

Activity Recognition in Multi-User Environments Using Techniques of Multi-label Classification

Activity recognition represents the cornerstone in realizing intelligent services such as energy conservation and ambient assisted living in smart environments. The problem statement of most activity recognition research assumes that only mutually exclusive activities occur in smart environments. The majority of research projects in this field focus on single-user environments where only one user performs a single activity at a given time. Such solutions are not applicable in real-world scenarios where multiple users reside in a home performing co-temporal activities. Our work addresses the problem of activity recognition in multi-user environments by utilizing the techniques of multi-label classification. It is based on a multi-label activity recognition dataset which we collected by deploying appliance-level power sensors as well as environmental sensors in a two-person apartment. In this dataset, a feature vector of sensor readings can have more than one label indicating the occurrence of more than one activity at a given time. In this work, we show that recognizing activities in smart environments can be achieved solely based on fine-granular power consumption data and without the need for installing any other sensing modality. Moreover, we prove that extracting and utilizing dependency relations between concurrent activities as well as temporal relations between subsequent activities provide a crucial enhancement of the predictive accuracy of activity recognition models.

Electricity-Metering in a Connected World: Virtual Sensors for Estimating the Electricity Consumption of IoT Appliances

Due to rising electricity prices, there is an increasing incentive to save energy. Therefore, more and more large organizations intend to reduce their energy consumption. Often, their plans cannot be realized due to missing insights into the causes energy consumption. Centralized energy meters provide no information at which appliances the energy is spent and the installation of thousands of distributed meters is often not feasible from an economic point of view. To simplify the energy metering in large scale, we propose to make Internet of Things (IoT) appliances aware of their own electricity consumption using on software based virtual energy sensors. We demonstrate how to automatically generate those energy models for nearly arbitrary networked devices with a high accuracy. Our purely software based energy metering solution approximates the energy consumption of common office equipment with an error between 2.19% and 10.8%. Using our approach, IoT appliances become aware of their own energy expenditure. This greatly simplifies energy metering on device level granularity, giving appropriate user feedback and developing more energy-efficient appliances. All these benefits are achieved without the need for installing additional hardware sensors.

Hybrid Models for Short-Term Load Forecasting Using Clustering and Time Series

Short-term forecasting models on the micro-grid level help guaranteeing the cost-effective dispatch of available resources and maintaining shortfalls and surpluses to a minimum in the spot market. In this paper, we introduce two time series models for forecasting the day-ahead total power consumption and the fine-granular 24-hour consumption pattern of individual buildings. The proposed model for predicting the consumption pattern outperforms the state-of-the-art algorithm of Pattern Sequence-based Forecasting (PSF). Our analysis reveals that the clustering of individual buildings based on their seasonal, weekly, and daily patterns of power consumption improves the prediction accuracy and increases the time efficiency by reducing the search space.

Hybrid Model for Large Scale Forecasting of Power Consumption

After the electricity liberalization in Europe, the electricity market moved to a more competitive supply market with higher efficiency in power production. As a result of this competitiveness, accurate models for forecasting long-term power consumption become essential for electric utilities as they help operating and planning of the utility’s facilities including Transmission and Distribution (T&D) equipments. In this paper, we develop a multi-step statistical analysis approach to interpret the correlation between power consumption of residential as well as industrial buildings and its main potential driving factors using the dataset of the Irish Commission for Energy Regulation (CER). In addition we design a hybrid model for forecasting long-term daily power consumption on the scale of portfolio of buildings using the models of conditional inference trees and linear regression. Based on an extensive evaluation study, our model outperforms two robust machine learning algorithms, namely random forests (RF) and conditional inference tree (ctree) algorithms in terms of time efficiency and prediction accuracy for individual buildings as well as for a portfolio of buildings. The proposed model reveals that dividing buildings in homogeneous groups, based on their characteristics and inhabitants demographics, can increase the prediction accuracy and improve the time efficiency.

Evaluation of user feedback in smart home for situational context identification

In the recent years, smart home projects started to gain great attention from academic as well as industrial communities. However, an essential challenge that all smart home ideas face is the provision of the ground truth i.e. the labeled training data required to train the machine learning algorithms which achieve the smartness of the smart home. Another challenging task is to evaluate the correctness of the collected ground truth so that we can be sure that we train the system with correct data which represents the reality. In order to build a smart home which is interactive and adaptable to the behavior and preferences of its inhabitants, we need to have comprehensive information about the everyday behavior and preferences of the inhabitants of the smart home. This comprehensive information which needs to be collected represents the ground truth in the context of our smart home research. Many technologies have been utilized in order to collect this information. In this paper, we present our approach for collecting the ground truth in smart homes in a nonintrusive way. More importantly, we present our methodology for evaluating the correctness of the collected ground truth.

Enhancing user privacy by data driven selection mechanisms for finding transmission-relevant data samples in energy recommender systems

In order to find energy saving potentials, future home energy recommender systems needs a large database of historic energy consumption information from various appliances. Having reference data, those systems could decide whether an appliance is wasting energy or not. However, the collection of this reference data degrades the user privacy as energy traces contain sensitive information which allows the exhibition of user behavior. In order to mitigate those privacy implications, we propose a method of sparse data collection. Our proposed solution minimizes the amount of collected reference data by removing energy traces which do not provide new information for the recommender system. Our proposed solution is capable of reducing the collected amount of data by a factor of 2 without lowering the accuracy of the future home energy recommender system.