User in the Loop: Adaptive Smart Homes Exploiting User Feedback—State of the Art and Future Directions

Abstract

Due to the decrease of sensor and actuator prices and their ease of installation, smart homes and smart environments are more and more exploited in automation and health applications. In these applications, activity recognition has an important place. This article presents a general architecture that is responsible for adapting automation for the different users of the smart home while recognizing their activities. For that, semi-supervised learning algorithms and Markov-based models are used to determine the preferences of the user considering a combination of: (1) observations of the data that have been acquired since the start of the experiment and (2) feedback of the users on decisions that have been taken by the automation. We present preliminarily simulated experimental results regarding the determination of preferences for a user.

A Review of the Recent Developments in Integrating Machine Learning Models with Sensor Devices in the Smart Buildings Sector with a View to Attaining Enhanced Sensing, Energy Efficiency, and Optimal Building Management

“…Analyzing the papers selected and summarized in Table S13, it can be observed that 78% of them exclusively analyze smart homes, 16% take into consideration smart buildings in general, 3% analyze both smart homes and buildings, while the remaining 3% of the selected papers refer to smart workplace environments. The authors of these scientific articles make use of different types of sensors in their analyses, including wireless sensor networks [70,115,[119][120][121][122][123][124]; passive infrared motion sensors [82,97,117,118,122,125,126]; motion sensors [25,70,81,118,120,127,128]; environmental sensors [10,25,81,82,[116][117][118]123,[127][128][129][130][131][132]; temperature sensors [116,118,120,123,125,[131][132][133]; humidity sensors…”

Arcades: A deep model for adaptive decision making in voice controlled smart-home

“…On the other hand, the comfort cost, which is induced when the inhabitant have to modify the decision made by the system. Although this project showed promising results, it seems to be no longer active but has given rise to a number of researches by other groups [60,23] which showed that RL allows some flexibility that traditional machine learning paradigms do not. Nevertheless, if RL is particularly suitable for restricted environment (with a few possible states and actions), it poorly scales to large real cases sets.…”

System Design for Coordinated Multi-robot Assistance Deployment in Smart Spaces

“…In previous work, semisupervised learning and Markov-based models were used to generate adaptive and personalized decisions in multi-users environments [11], [12]. In this work, we consider that users through their reactions and feedback (implicit or explicit) are an important source of information on how to achieve tasks in an adaptive and personalized way (time to give medicine, path recommendation, etc.)…”