[Paper Review] Internet of Behavior (IoB) and Explainable AI Systems for Influencing IoT Behavior

Haya Elayan, Moayad Aloqaily+2 more|arXiv (Cornell University)|Sep 15, 2021

Internet of Things and AI15 citations

TL;DR

This paper proposes an Internet of Behavior (IoB) and Explainable AI (XAI)-integrated system to influence household IoT energy consumption behavior. By leveraging deep learning for power consumption prediction and XAI for model transparency, the system reduces active power usage by 522.2 kW and saves €95.04 in costs over 200 hours, demonstrating a viable path to energy sustainability through behavior-driven IoT control.

ABSTRACT

Pandemics and natural disasters over the years have changed the behavior of people, which has had a tremendous impact on all life aspects. With the technologies available in each era, governments, organizations, and companies have used these technologies to track, control, and influence the behavior of individuals for a benefit. Nowadays, the use of the Internet of Things (IoT), cloud computing, and artificial intelligence (AI) have made it easier to track and change the behavior of users through changing IoT behavior. This article introduces and discusses the concept of the Internet of Behavior (IoB) and its integration with Explainable AI (XAI) techniques to provide trusted and evident experience in the process of changing IoT behavior to ultimately improving users' behavior. Therefore, a system based on IoB and XAI has been proposed in a use case scenario of electrical power consumption that aims to influence user consuming behavior to reduce power consumption and cost. The scenario results showed a decrease of 522.2 kW of active power when compared to original consumption over a 200-hours period. It also showed a total power cost saving of 95.04 Euro for the same period. Moreover, decreasing the global active power will reduce the power intensity through the positive correlation.

Motivation & Objective

• To investigate the concept of the Internet of Behavior (IoB), including its workflow, benefits, challenges, and industrial applications.
• To design a trusted and comprehensible system that integrates IoB with Explainable AI (XAI) to influence and modify IoT-based user behavior.
• To evaluate the system in a real-world use case involving household electrical power consumption to reduce energy waste and cost.
• To analyze the impact of reduced global active power on global energy intensity, leveraging their strong positive correlation.
• To identify future research directions, including user feedback integration, distributed system design, and enhanced security and privacy mechanisms.

Proposed method

• Employing IoT sensors to collect real-time household electrical power consumption data for training and inference.
• Using deep learning models to predict global active power for the next hour based on historical data from the past four years.
• Implementing XAI techniques to provide interpretable explanations for AI predictions, enhancing user trust and system transparency.
• Applying a threshold-based control mechanism: if predicted power exceeds the historical average for the same hour, weekday, and month, the system triggers power reduction actions.
• Validating the system on 200 test samples to measure energy savings, cost reduction, and impact on global intensity.
• Utilizing a positive correlation between global active power and global intensity to infer broader environmental benefits from reduced power usage.

Experimental results

Research questions

• RQ1 How can IoB and XAI be integrated to create a trustworthy and transparent system for influencing IoT-based user behavior?
• RQ2 To what extent can an IoB-XAI system reduce household electrical power consumption and associated costs in a real-world scenario?
• RQ3 What is the impact of reduced global active power on global energy intensity, given their strong positive correlation?
• RQ4 How can user feedback and instructional notifications improve long-term behavior change and system engagement?
• RQ5 What are the technical and experiential trade-offs of transitioning from a centralized to a distributed system architecture?

Key findings

• The proposed IoB-XAI system reduced global active power by 522.2 kilowatts over a 200-hour testing period.
• Total cost savings amounted to €95.04 over the same 200-hour period, based on a household electricity rate of €0.182 per kilowatt-hour.
• The system issued warnings on 27 out of 200 predicted hours where consumption exceeded historical averages, triggering power-saving actions.
• A strong positive correlation (99%) between global active power and global intensity means that reducing active power also reduces global intensity.
• The system’s use of XAI enhances user trust by providing interpretable explanations for AI-driven decisions.
• Future enhancements such as user feedback integration, distributed architecture, and advanced security measures are expected to further improve system performance and adoption.